Ensuring the Best Start in Life

Targeting versus Universality in Early Childhood Development

Gillian Doherty 6 décembre 2007

Dans le cadre du débat sur la nécessité d’adopter un système de services de garde et d’éducation de la petite enfance, on avance souvent qu’il faut plutoÌ‚t affecter les ressources disponibles aux enfants qui en ont le plus besoin. Or nous en savons très peu sur l’ef- ficacité des initiatives ciblées actuellement en place au Canada et auxquelles sont alloués plus de 260 millions de dollars par année. Dans cette étude, Gillian Doherty analyse les effets de ces programmes et se demande s’ils produisent le rendement social escompté par rapport aux investissements publics qui leur sont consacrés. Ce qui l’amène à s’interroger plus généralement sur la meilleure des deux approches, universelle ou ciblée, pour assurer aux enfants le meilleur départ possible dans la vie.

Elle passe tout d’abord en revue les programmes d’inter- vention précoce destinés aux enfants vulnérables susceptibles de vivre des problèmes de développement et ce que nous savons de leur incidence sur le développement des enfants. Elle examine 13 initiatives canadiennes de trois types : cen- trées sur les parents, sur les enfants et sur les deux généra- tions. Elle analyse pour chacune la population cible, les services fournis, la durée de l’intervention, le nombre d’en- fants/familles bénéficiaires, la formation du personnel, de meÌ‚me que les sommes investis par les gouvernements.

Elle examine également l’efficacité relative de ces pro- grammes en ce qui concerne le développement des enfants et, à la lumière des résultats tirés de recherche menées au Canada et aux États-Unis, elle conclut que :

  • Les initiatives qui ciblent directement les enfants et qui sont offerts en institution avec un programme struc- turé produisent les effets les plus positifs sur le développement des enfants vulnérables.

  • Bien que les interventions centrées sur les parents avec l’objectif d’améliorer les aptitudes parentales, la forma- tion ou l’employabilité puissent profiter aux parents, notamment en raffermissant leur confiance en soi, ils ont des effets négligeables sur le développement des enfants.

  • L’efficacité des programmes pour enfants dépend de leur qualité, c’est-à-dire qu’il sont donné par un per- sonnel bien formé et dispose d’une programmation et d’un ratio éducateur/enfant adéquats.

  • L’efficacité des programmes dépend également de leur durée.

L’auteure note que les programmes de soutien aux enfants vulnérables sont peu nombreux au Canada et qu’ils ne touchent qu’un petit nombre des enfants visés. En fait, comme la plupart de ces programmes ne font l’ob- jet d’aucune évaluation ou suivi, il est impossible d’établir dans quelle mesure ils atteignent leur clientèle cible. Qui plus est, 60 p. 100 des sommes investies dans ces programmes sert au financement d’initiatives qui ont une incidence négligeable sur le développement des enfants.

L’auteure examine ensuite les données internationales sur les avantages d’une approche universelle pour aider les enfants vulnérables. Elle conclut que, malgré son couÌ‚t ini- tial plus élevé, l’approche universelle est plus prometteuse que l’approche ciblée pour deux grandes raisons.

Premièrement, elle permet de rejoindre un plus grand nombre d’enfants vulnérables. Doherty rappelle que si le risque de connaiÌ‚tre des problèmes de développement est plus élevé chez les enfants vivant dans les familles les plus dému- nies, de récentes enqueÌ‚tes ont révélé qu’il existe des enfants vulnérables dans tous les groupes socio-économiques et que plus de 70 p. 100 des enfants vulnérables au Canada sont en fait issus de familles qui ne vivent pas dans la pauvreté. Ceci signifie que les programmes ciblés sur les enfants issus des familles les plus pauvres revient à priver de soutien la majorité des enfants qui éprouvent des difficultés. L’expé- rience de plusieurs pays montre que des programmes d’édu- cation pour la petite enfance abordables et accessibles suscitent une forte participation ce qui, du meÌ‚me coup, per- met de rejoindre un plus grand nombre d’enfants qui ne vivent pas dans des familles pauvres.

Deuxièmement, les données internationales indiquent que la maturité scolaire des enfants de tous les milieux socio-économiques et leurs chances de réussite à long terme se trouvent améliorés par des programmes universels de qualité, que ce soit en prématernelle ou en garderie. Elle examine finalement les analyses couÌ‚t-bénéfices des initia- tives ciblées et des programmes universels, pour en con- clure que les deux approches peuvent produire pour la société des avantages supérieurs à leurs couÌ‚ts.

En terminant, l’auteure analyse les implications de ces résultats pour les politiques canadiennes en matière de petite enfance. Puisque qu’environ 25 p. 100 des enfants canadiens de cinq ans manquent de maturité scolaire, que les programmes ciblés sont relativement inefficaces et que l’accès et la qualité des services de garde à l’échelle du pays sont présentement fort limités, elle estime qu’il nous faut reconnaiÌ‚tre que (1) le risque de connaiÌ‚tre des problème de développement s’observe dans les familles de tous les niveaux de revenu ; (2) aucun repère observable n’est infail- lible pour ce qui est d’identifier tous les enfants vulnérables. Aussi, recommandent-elles que les services de garde et d’éducation de la petite enfance doivent eÌ‚tre universelle- ment accessibles à toutes les familles qui désirent en pro- fiter. AÌ€ son avis, il est impératif d’adopter une stratégie efficace, non ciblée, fondée sur les résultats de la recherche et qui dispose des ressources et du financement publics nécessaires à sa réussite. 

Introduction

Canada is facing a workforce shortage resulting from low birth rates and increasing numbers of baby boomers retiring over the next two decades. These demographics and the demands of new technolo- gies and the global economy for workers with good peo- ple, literacy, numeracy, problem-solving and decision-making skills mean that Canada’s future pros- perity depends upon increasing the productivity of what will be a much smaller workforce in the future. Increasing productivity requires public policies supported by adequate levels of government funding to ensure that every child reaches his or her fullest potential (Dodge 2003; Fortin 2006; Lefebvre and Merrigan 2003).

The research evidence is clear; the foundations for good adult physical health, social competence, communi- cation skills, adaptability, literacy, and numeracy are laid down before entry into kindergarten (McCain, Mustard and Shanker 2007). Children who enter school lacking the basic knowledge and skills required to benefit from the school program are more likely to experience social and academic difficulties and drop out prior to graduation (Heckman 2000; Stipek 2001). The costs of failure to graduate from high school are substantial for the individ- ual and for society. The individual faces decreased employability and earning potential and an increased likelihood of lifelong levels of poorer physical and mental health (Canadian Council on Learning 2006; Keating and Hertzman 1999). Society faces decreased workforce pro- ductivity, decreased government revenue and increased demands for social assistance and health services. For instance, adults with less than a high school diploma —19 percent of the population — contribute 9 percent of income taxes while consuming 35 percent of government income transfers such as employment insurance and social assistance (Canadian Council on Learning 2006).

Recognizing the need to give all Canadian children the best possible start in life, the First Ministers estab- lished the Federal-Provincial-Territorial Council of Ministers on Social Policy Renewal in 1996 to review existing social programs, engage in broad public con- sultation and develop a comprehensive strategy to support the development of young children. The result, known as the National Children’s Agenda (NCA), artic- ulates a vision of Canada as a country where all chil- dren have the opportunity to reach their full potential as adults and a policy framework for intergovernmen- tal initiatives to support young children and their fam- ilies1 (Federal-Provincial-Territorial Council of Ministers on Social Policy Renewal 1999). The NCA was followed in 2000 by the federal-provincial- territorial Early Childhood Development (ECD) Agreement under which the federal government com- mitted to transfer $500 million annually by 2007-08 to the provinces and territories to fund the promotion of healthy pregnancy, birth and infancy; parenting and family support; child development programs; and strengthening community supports (Canadian Intergovernmental Conference Secretariat 2000). In practice, only a small proportion of these funds was used for providing direct services to children.

The subsequent Multilateral Framework Agreement on Early Learning and Child Care announced in 2003 focused specifically on children’s programs. Through it, the federal government undertook to transfer $900 million to the provinces and territories over five years for services in settings such as child care programs and nursery schools (Human Resources Development Canada 2003). The 2004 Speech from the Throne sig- nalled the federal government’s wish to develop, in cooperation with the provinces and territories, a pan- Canadian early childhood education and care (ECEC) system based on the principles of quality, universality, accessibility and child development (Government of Canada 2004). In the following year, the federal budg- et committed $1 billion annually for five years to sup- port the development of such a system (Government of Canada 2005a). During the next few months, bilat- eral agreements were signed with each of the 10 provinces and initial federal funding transferred to them. These agreements and their related federal funding laid a foundation for the establishment of universal ECEC services across the whole country.

The election in 2005 resulted in a change of gov- ernment and the announcement that the recently signed bilateral ECEC agreements and their future funding would be terminated effective April 1, 2007.

As a replacement, the new government instituted the Universal Child Care Benefit (UCCB), a taxable pay- ment of $100 a month to all families for each child under age six, and allocated $250 million annually beginning in 2007 to create new child care spaces (Government of Canada 2006). The UCCB has been criticized as being insufficient to make regulated child care more affordable for most parents and fail- ing to address the need to expand child care avail- ability. Subsequently, the government promised in the March 2007 budget to provide additional annual funding of $250 million to provinces and territories for the creation of child care spaces (Government of Canada 2007a). The promise of this additional fund- ing in 2007, in combination with the $250 million announced in 2006, still represents a substantial cut to the $1 billion committed by the previous govern- ment in 2005.

During the discussions leading up to and following the NCA, there was little debate about targeted initia- tives for children deemed vulnerable to poor develop- mental outcomes or their role in an overall ECEC system. Yet such initiatives have been government- funded in Canada for over 15 years. By 1990, Nova Scotia was supporting 30 part-time programs based on the American Head Start program for preschoolers living in low-income families and Quebec was fund- ing maternelles quatre ans in low-income urban com- munities and maisons maternelles in targeted rural areas (Doherty 1991). In 1990, the Ontario govern- ment created Better Beginnings, Better Futures as a 25-year project to explore the potential of targeted programs to prevent developmental problems among children living in economically disadvantaged neigh- bourhoods (Peters et al. 2000). The federal govern- ment’s Community Action Program for Children (CAPC) began in 1993 (Boyle and Willms 2002). Subsequently, additional government money has been invested to continue and/or expand some of these programs and to fund new targeted initiatives such as the federal Aboriginal Head Start programs for on- and off-reserve children, Manitoba’s Families First program, Quebec’s Services intégrés en périnatalité et pour la petite enfance à l’intention des familles
vivant en contexte de vulnérabilité (SIPPE) and Saskatchewan’s KidsFirst program.

Currently, targeted ECEC initiatives receive in excess of $260 million from various sources, of which $145 million comes from the federal government and $99 million from provincial and territorial govern- ments (see table 7, page 28). Given the limited funding available for services to young children, it is important to examine what we know about the cost- effectiveness of targeted programs and engage in a debate about whether they provide the largest social return for the ECEC money they receive.

Objectives

This paper seeks to identify the most promising strategies for enhancing the development of vulnera- ble children and to contribute to the debate about how best to achieve the NCA goal of ensuring that all children in Canada have the best possible start in life.

It begins by describing Canadian early interven- tion programs for vulnerable children and what we know about their impact on children’s development. Since the body of Canadian evaluations is extremely small, Canadian findings are supplemented with American findings from similar initiatives. Concern is often and appropriately expressed about the validity of using American experience to inform Canadian policy development because of the differences between the two countries in factors such as the availability of universal health care and the extent of the social safety net. Nevertheless, it is the only option, given that few evaluations of early interven- tion strategies have been conducted in countries other than the United States. All Canadian evalua- tions are reported regardless of the rigour of their methodology and the extent to which it isolates the effect of the intervention from other influential fac- tors. However, only American evaluations with rigor- ous methodological designs that have been reported in a peer-reviewed journal or conducted by a research organization such as the National Institute for Early Education Research at Rutgers University are cited.

The paper then considers four specific policy issues: the relative efficiency of a targeted versus a universal approach for assisting vulnerable children; the conditions required for high-quality ECEC; the benefit/cost ratios associated with different types of early childhood interventions; and what is required to support the healthy development of all Canada’s chil- dren. In conclusion, the paper identifies a need to rethink the current targeted approach to enhancing the development of vulnerable children.

Scope

Willms uses the term “vulnerable children” to denote young children susceptible to poor developmental out- comes as a result of their environmental circumstances and states that this does not include children whose susceptibility is associated with physical or mental dis- abilities or health problems (2002). Consistent with Willms’s definition, this paper focuses on interventions for children prior to school entry who had a normal birth weight and do not exhibit indications of a disability or evidence of conditions such as fetal alcohol syndrome (FAS) but do have developmental problems. Thus, the paper does not include discussion of the Canada Prenatal Nutrition Program (CPNP) and similar provincial or terri- torial targeted initiatives to improve the nutrition of the fetus or services for special needs children.

Two broad community development initiatives tar- geting areas where children are deemed vulnerable because their neighbourhood has a high concentration of low-income families or families of Aboriginal ances- try are excluded. The first — 1,2,3 Go! — does not pro- vide any direct interventions but instead responds to issues of concern to the community by mobilizing com- munity members to work together to address them (Centraide of Greater Montreal 2007). The second, Brighter Futures for First Nations and Inuit Communities, provides funds that may be used for a wide variety of purposes such as community mental health promotion; public awareness and prevention activities related to family violence or suicide; improv- ing birth outcomes; parent education; and/or cultural activities (Government of Canada 2007b). While some of these activities directly or indirectly address children’s vulnerability to developmental problems, the broad mandate and emphasis on culturally appropriate approaches has resulted in initiatives that are very com- munity-specific. This makes it difficult to tease out which strategies are appropriate for replication in other communities. However, CAPC, another broad communi- ty initiative, is discussed. CAPC funding is specifically tied to services for children under age six and their families and is intended to enhance the children’s devel- opment (Government of Canada 2007b) Almost every site provides direct services to children and/or parenting education and supports parents in their parenting role (Beaudoin and Turcotte 2002).

The National Longitudinal Survey of Children and Youth (NLSCY), the Étude longitudinale du développe- ment des enfants du Québec (ELDEQ) and the federal government’s Aboriginal Children’s Survey (ACS) sup- port and inform the development and implementation of initiatives for vulnerable children by identifying how child, family and community factors influence child development and expanding our understanding of child vulnerability. Other initiatives, such as the federal gov- ernment’s Understanding the Early Years Project and the Early Child Development Mapping Project being undertaken by the Human Early Learning Partnership in British Columbia, support and inform initiatives in specific communities. Their provision of information about a community’s socio-economic and demo- graphic characteristics, the availability and location of resources within it, and the developmental level of its child population assist communities to address resource gaps and monitor the effect of actions taken. Neither longitudinal surveys nor community mapping engage in direct service provision and further discus- sion of them is outside the scope of this paper. Additional information about these initiatives and their findings can be found in Kershaw et al. (2005) and McCain, Mustard and Shanker (2007).

Identification of targeted initiatives

Canadian initiatives to enhance vulnerable children’s development directly by working with the children and/or indirectly through parent education were identified through the federal government’s report on its early childhood activities (Government of Canada 2007b); federal, provincial and territorial government Web sites; and leads provided by individuals. Specific information about program design and child out- comes was collected from material posted on Web sites, documents provided by the programs, published research and data obtained through personal commu- nication. In order to keep the length of the report rea- sonable, the discussion of American initiatives for vulnerable children and their outcomes is limited to initiatives similar to those existing in Canada. Comparable American initiatives were identified through the comprehensive reviews undertaken by Gormley (2006) and Karoly, Kilburn and Cannon (2005); these also provided some information about the outcomes of these initiatives. Additional informa- tion on child outcomes was sought through a key- word search of the University of Toronto Library catalogue. The findings from single-site demonstra- tion initiatives such as the Carolina Abecedarian and the High/Scope Perry Preschool projects are not included in the report because they were extremely expensive; for example, it cost $15,270 per child per school year in 2003 dollars for the Perry Preschool Project (Aos et al. 2004). Such expenditures would not be feasible to replicate on a large-scale basis. Information on the outcomes of demonstration proj- ects such as the High/Scope Perry Preschool is sum- marized in Anderson et al. (2003), Aos et al. (2004) and Karoly, Kilburn and Cannon (2005).

Targeted Initiatives: What Do We Know?

Arobust body of Canadian research documents that children living in a low-income family, a family that is Aboriginal, and/or a family whose home language is other than that used in the community at large are less likely to enter school with the skills required to take advantage of what it offers (Doherty 2007). Each of these three family characteristics can be viewed as a marker indicating children’s increased vulnerability to poor develop- mental outcomes. The existence of such easily identi- fiable markers makes a persuasive case for targeting the provision of initiatives intended to enhance young children’s development to communities with relatively large proportions of such families.

Targeted initiatives to promote the development of children deemed vulnerable to poor developmental outcomes exist in all parts of Canada. They may be fully funded by the federal and/or a provincial gov- ernment or rely on a combination of government funds, donations and their own fundraising. The types of activities engaged in depend upon the prem- ise on which the initiative is based. When the basic premise is that children’s development is best enhanced by working directly with them, the primary focus is on group programming for children. Initiatives that focus on providing parenting educa- tion and support are based on the premise that the child’s development can be best enhanced by chang- ing parenting behaviour and the home environment.2 Other initiatives operate on the premise that changing vulnerable children’s developmental trajectories requires working with both the parent and the child and, to a greater or lesser extent, these initiatives provide both programs for children and education and support for parents.

In an effort to provide information in the most useful way for policy development, commentators in both Canada and the United States have adopted the convention of discussing initiatives operating under the same premise as a single group, using a three- part categorization scheme of parent-focused, child- focused and two-generation programs. In reality, not all initiatives fit neatly into this categorization; for example, Head Start focuses on children’s program- ming but also provides some parent supports includ- ing assisting parents to access other community resources. Nevertheless, the three-part categorization scheme is a useful policy tool in that it enables con- sideration of the relative effectiveness of focusing on children, focusing on parents, or providing an inter- vention that combines parent and child services.

This section provides an overview of each category of intervention, beginning with parent-focused initia- tives and concluding with two-generation programs. Each of the three categories is organized as follows: (1) Canadian initiatives; (2) evaluations of Canadian initiatives; (3) evaluations of comparable American initiatives; and (4) discussion. A summary table is provided for each of the Canadian initiatives in each category, indicating the target population, services provided, duration of intervention and approach to staffing. Table 7 (page 28), provides the best available information on federal and provincial/territorial funding for targeted initiatives to promote the devel- opment of vulnerable children and the number of children/families receiving assistance.

Parent-focused initiatives

Parent-focused initiatives seek to enhance the child’s development indirectly by changing parental behav- iour and enhancing the home environment; they pro- vide parents with information about child development and effective parenting strategies and teach them how to engage in educational activities with their children. The service is usually delivered through home visits but may be provided through group or individual par- ent meetings in an agency’s office. Home visiting has the advantage of enabling linguistic and cultural matching between staff and family, may be less threat- ening for some parents, and eliminates the need for the parent to travel in order to receive the service. The home visitor focuses on the parent during the home visit; direct contact between visitor and child is limited to modelling how an activity should be done.

There are five Canadian parent-focused initiatives: Manitoba’s Families First program, Quebec’s Services intégrés en périnatalité et pour la petite enfance à l’intention des familles vivant en contexte de vul- nérabilité (SIPPE), Saskatchewan’s KidsFirst program, the Home Instruction for Parents of Preschool Youngsters (HIPPY) Program and the Parent-Child Mother Goose Program. Table 1 provides basic infor- mation about each of these five initiatives.

As illustrated in the table, all the initiatives pro- vide parents with information on child development, effective parenting and activities to stimulate chil- dren’s development and all but Mother Goose deliver their interventions primarily through home visits.

However, they vary on two other important dimen- sions: the intensity of the intervention and the back- ground of the person who delivers the program. At one end of the intensity continuum, Quebec’s SIPPE and Saskatchewan’s KidsFirst initiatives begin working with the mother when she is pregnant and parents may con- tinue in the program until the child is five; they also provide assistance that goes beyond parent education. Mother Goose and HIPPY are at the other end of the continuum: both focus solely on teaching parents how to stimulate their child’s development through either weekly group meetings or biweekly home visits for a 30-week period. The Manitoba, Saskatchewan and HIPPY programs rely on paraprofessionals for service delivery while the programs operated by SIPPE and Mother Goose are staffed by people who have a profes- sional degree.

Manitoba’s Families First program

Families First, part of the provincial Healthy Child Manitoba initiative created in 2000, operates in each of the regional health authorities in the province and is delivered by them. As of March 31, 2006, the program had the capacity to provide home visits to 1,581 fami- lies using a 2005-06 budget of $9,486,000 (Government of Manitoba 2006; Darlene Girard, Healthy Child Manitoba, personal communication, March 20, 2007). It is important to note that this budget includes funding for the province’s universal screening of every family with a newborn in Manitoba as well as an in-depth assessment of families identified as possibly vulnerable to determine whether they should be referred to Families First.

In addition to providing assistance in accessing other community services, Families First uses a home-visiting program to deliver a specific parenting and child devel- opment curriculum that addresses basic care, health and safety; child development; parenting issues such as dis- cipline; and strategies to enhance family functioning. Home visitors create an individual binder of activities and information that is left with the family and discuss with parents ways to incorporate their new learning into their daily routine. There is no formal requirement that parents work with the materials between visits and families choose the level, intensity and duration of involvement. As indicated in table 1, home visitors receive preservice training through the Manitoba Curriculum for Training Home Visitors; this includes modules in child development and parenting, safety and well-being and the materials used in the program (Creating Great Kids and Growing Great Families). Over the subsequent 18 months, they take additional train- ing in areas such as domestic violence (Darlene Girard, personal communication, March 20, 2007).

Table 1 Descriptions of Targeted Parent Focused Initiatives

Quebec’s Services intégrés en périnatalité et pour la petite enfance à l’intention des familles vivant en contexte de vulnérabilité (SIPPE) SIPPE, which targets women who are pregnant and under age 20 and/or living in extreme poverty and/or are recent immigrants, is provided right across the province through regional Centres de santé et des serv- ices sociaux. The mother and the father may continue to participate in the program until the child is five (Quebec, Ministère de la Santé et des Services sociaux 2004). Participants are referred by hospitals, other health services and social services. In 2006-07, SIPPE served 5,240 women, an estimated 56 percent of the eligible population, and had an annual budget of $48 million (Louise Therrien, Quebec, Ministère de la Santé et des Services sociaux, personal communication, September 12, 2007).

Participants are provided with a variety of supports including home visits lasting 60 to 90 minutes every second week beginning in the twelfth week of preg- nancy, information about good nutrition and health practices, food coupons and prenatal vitamins. After the birth, there are weekly home visits until the child is six weeks old, then every second week until age twelve months and monthly until age five. The home visitor’s activities are tailored to the needs and aspira- tions identified by the family and may include the provision of information about child development, child nutrition, effective parenting strategies and edu- cational activities to do with the child; budgeting and other life-skill counselling; assistance in accessing other services such as child care and job training; and accompanying a parent to an appointment. Generally, the home visitor is a nurse, although sometimes the individual may have another post-secondary profes- sional credential, for example, in social work. The ministry has designed materials to help the regional centres provide in-service training. As an employee of a Centre de santé et des services sociaux, the home visitor has access to consultation with and can refer a client to a broad interdisciplinary team including nutritionists, social workers, psychologists, speech and language therapists, and medical doctors. Home visits are supplemented by group activities for parents and their children and parents are encouraged and assisted to enrol their preschool-aged child in regulated child care as a means of providing a group educational experience. An evaluation of SIPPE was started in 2005 and is ongoing (Quebec, Ministère de la Santé et des Services sociaux 2004; Louise Therrien, personal com- munication, September 12, 2007).

Saskatchewan’s KidsFirst program

Saskatchewan’s KidsFirst program, initiated by the gov- ernment in 2001, operates in the nine communities across the province deemed to have the greatest con- centration of vulnerable families. It is delivered by community agencies operating under the supervision of a KidsFirst Management Committee, with either a local school division or health region being the partner responsible to the province. At the end of March 2006, home visits were being made to 1,150 families and the program maintained 128 dedicated regulated child care spaces for its clients (Gail Russell, Early Learning and Child Care Branch, personal communication, March 12, 2007). The 2006-07 budget for KidsFirst was $14,033,000 (Government of Saskatchewan 2006). It is important to note that the budget includes not only the services listed in table 1 but also $2,185,000 granted to community child care services across the province to improve their program quality (Government of Saskatchewan 2006).

KidsFirst includes the provision of nutritional sup- plements and prenatal education to vulnerable preg- nant women; home visiting after the birth of the child with the delivery of a specific parenting and child development curriculum that addresses basic care, health and safety, child development and parenting issues such as discipline; dedicated mental health and addictions services for parents; and assistance in accessing other community resources such as literacy training. Initially, home visits occur once a week and then gradually decrease as the family’s capabilities progress; parents are expected to work with their chil- dren between home visits using the suggestions and materials provided by the home visitor.

There are also 128 regulated child care spaces oper- ated by other organizations in the community dedicat- ed for the use of KidsFirst parents to support them while they attend school, take skills upgrading, seek work or require respite. A space may also be used to provide an enhanced developmental opportunity for a child. KidsFirst may cover the whole or part of the child care fee, depending on the family’s situation. The mix of services received by a family is tailored on a case-by-case basis to provide only those supports need- ed to address its needs, with the goal that the family will be linked with services and support networks within the community and develop the necessary level of self-sufficiency and stability to leave the pro- gram. As indicated in table 1, home visitors are typi- cally paraprofessionals who have raised their own children and are given training related to the pro- gram (Government of Saskatchewan 2005; Gail Russell, personal communication, March 12, 2007).

Home Instruction for Parents of Preschool Youngsters (HIPPY)

HIPPY Canada is a nonprofit organization that receives its revenue from the federal government, donations and its own fundraising. In 2005-06, 385 low-income families were served, for an expenditure of $206,299 through six sites in British Columbia, one site in Montreal and one site in Toronto (HIPPY Canada 2006). Five of the HIPPY sites in British Columbia serve Aboriginal families. One of these sites is operated by an Aboriginal consortium and serves five First Nations, three other sites each serve a First Nation on-reserve community, and the remaining site serves Aboriginal families living in Vancouver (Debbie Bell, Executive Director, HIPPY Canada, per- sonal communication, March 12, 2007).

Children enter the program at age three and fami- lies are required to make a two-year commitment to participate in the program for 30 weeks during the school year. The HIPPY program is very structured and is implemented across sites in the same way using a standard curriculum and standard materials. These include HIPPY storybooks for parents to read with their children and materials designed to assist children develop visual-motor, language, discrimination and problem-solving skills. All of the parent instructional materials are prepared at a grade three level and are available in a number of languages. HIPPY Canada has modified the standard HIPPY program for its Aboriginal sites by replacing some of its storybooks with First Nations titles and developing a language enrichment program to support the transmission of the first language of the communities involved. HIPPY Canada has also contracted with Let’s Talk Science, a national organization that develops learning programs to improve science literacy, for the development of five different age-appropriate science activities to be included in the HIPPY standard age four curriculum in all its programs (HIPPY Canada 2006).

The family is visited at home every second week by a HIPPY home visitor who brings a packet of materials to be used in activities with the child and spends about an hour reviewing the materials with the parent, using role-playing. Between visits, parents are expected to spend a minimum of 15 minutes a day doing activities with their child (HIPPY Canada undated). There is a group meeting at a community centre or school every other week where parents may discuss their progress, ask questions or seek advice. The home visitors are parents from the community who have participated in HIPPY and receive preser- vice training related to the implementation of the HIPPY program and ongoing weekly training sessions for an average of 140 training hours in their first year (HIPPY Canada 2006; Debbie Bell, personal com- munication, March 12, 2007).

The Parent-Child Mother Goose program

The Mother Goose Program (Programme la Mère l’Oie pour parents et enfants) is a nonprofit organization that operates in eight provinces and the Yukon. It receives funding through provincial and municipal government grants, fees that agencies pay to obtain training for their staff and be permitted to use the Mother Goose programming materials and overall approach, and donations. Two versions of Mother Goose are available: one targets children under age two; the other, children between the ages of two and four. Each version provides a one-hour group experi- ence for parents/caregivers and their children for 30 consecutive weeks.

The actual program is delivered by staff from agencies working with vulnerable children in an out-of-home setting such as the agency’s office rather than by Mother Goose employees. There is not a specific curriculum, such as exists in HIPPY. Instead, parents/caregivers receive demonstrations and education on how to use interactive rhymes, stories and songs with their child to enhance the child’s language and communication skills. Printed versions of the rhymes and songs are supplied to participants for use at home but there is no requirement that parents work with their child between sessions. The person leading the program is required to have participated in a two-day Mother Goose training workshop and obtained hands-on experience in a mandatory practicum under the supervision of a fully trained person for a minimum of 30 sessions. Additional training workshops are available for staff members who wish to participate in them (Parent-Child Mother Goose Program 2007; Dorota Dziong, Parent-Child Mother Goose Program, personal communication, September 10, 2007).

In 2005-06, the Mother Goose national organiza- tion and its largest provincial organization, in Toronto, had a combined budget of $27,595. Mother Goose organizations operating in other jurisdictions and/or the agencies delivering the program may obtain additional funds for this purpose. Nationally, the agencies delivering the program reported having served a total of 1,649 adults (Dorota Dziong, person- al communication, September 25, 2007).

Evaluations of Canadian parent-focused initiatives

The three Canadian parent-focused initiatives that have been evaluated are Families First in Manitoba, KidsFirst in Saskatchewan and one of the HIPPY sites in British Columbia. Manitoba and Saskatchewan both use data obtained from all the families that par- ticipated in the NLSCY in 1998-99, 2000-01 and 2002-03 as performance indicators for their parent- focused programs (Government of Manitoba 2006; Government of Saskatchewan 2005).

Manitoba reports changes in parental behaviour and family function across the period but no infor- mation about changes in children’s development lev- els. However, American research indicates that successfully influencing parenting style or the fre- quency of parents’ engagement in educational activi- ties with their children is not always associated with enhanced child development (Gomby 2005). Saskatchewan reports reductions in the percentage of behaviour problems and improvement in average scores measuring cognitive development for children in the total sample of NLSCY families in the province between 1998-99 and 2002-03. However, it is diffi- cult to know what this means without knowing what the changes were among children in the subsample of NLSCY families with similar demographic and socio- economic characteristics to those of the participants in the intervention program. Saskatchewan has recently obtained a grant from the Canadian Population Health Institute for a three-year evalua- tion of its KidsFirst program (Gail Russell, Saskatchewan Early Learning and Child Care Branch, personal communication, March 12, 2007). This eval- uation should enable the provision of more specific information.

A quasi-experimental evaluation has been con- ducted in a single HIPPY site in British Columbia involving 14 children whose parents participated in HIPPY for two years and two comparison groups of children who were individually matched with the HIPPY children on gender, ethnicity and family socio- demographic variables but whose parents had not par- ticipated in HIPPY (LeMare and Audet undated).3 All the children in one comparison group had attended a centre-based preschool program while the 14 children in the other comparison group did not have any pre- school experience. At the end of the kindergarten year, the HIPPY children outperformed the children in the comparison groups on standard measures of knowledge of concepts, overall cognitive development, and degree of comfort with the school situation and were rated higher by their kindergarten teacher on a standard measure of school adjustment and peer relationship skills. This pattern of results is positive; however, none of the differences was statistically significant. The lack of significance may reflect the small sample size, a fac- tor that decreases the likelihood that statistical analyses will detect even moderate-sized effects (Karoly, Kilburn and Cannon 2005).

Evaluations of comparable American parent- focused initiatives

Quebec’s SIPPE and Saskatchewan’s KidsFirst initiatives appear to owe much of their design to the Nurse-Family Partnership (NFP) program pioneered in Elmira, New York, by David Olds and his colleagues over 20 years ago. The initial program targeted pregnant women with no previous live births who met one of the following eligibility criteria: under age 19; unmarried or living in poverty. The women received home visits by public health nurses lasting 75 to 90 minutes starting in the prenatal period and continuing until the child’s second birthday. The frequency of the visits varied according to the needs of the mother, with a mean of 9 during preg- nancy and 23 after the child’s birth. The intent was to improve the women’s prenatal health, and thus birth outcomes, reduce the incidence of neglect and abuse of the children, and improve the women’s life situations. The nurses followed detailed visit-by-visit protocols, implemented specific interventions and assisted partici- pants to access other health and human services (Olds et al. 1998; Olds et al. 1999). According to the NFP Web site, the program is now operating in 23 states.

The original Elmira program has been subject to a longitudinal evaluation and also replicated and evalu- ated through randomized trials in Memphis, Tennessee, and Denver, Colorado.4 Fifteen years after entry into the Elmira program, participants who had been unmar- ried and living in poverty had longer gaps between the births of subsequent children, received welfare for sig- nificantly fewer months and had fewer arrests by the police and fewer incidents of verified child neglect or abuse (Olds et al. 1998; Olds et al. 1999). The Memphis and Denver sites also report that partici- pants had longer gaps between births. Memphis also found lower use of welfare than found in the com- parison group two years after leaving the program, although Denver did not find any between-group dif- ferences on this variable (Olds et al. 1999, 2004b).

The NFP’s original goals did not include a focus on enhancing child development or school-readiness and the results in these areas vary across sites and among families, with different characteristics within a site. Results are available for children up to age four for Denver, age six for Memphis and age fifteen for Elmira. At age four, children in the Denver site whose parents had received home visits by nurses obtained higher scores on a language test and on a test measuring their capacity for focused attention and self-regulation than did children in the comparison group (Olds et al. 2004b). However, four-year-old participants in the Elmira pro- gram had no better ability to remain focused or to self- regulate than the comparison children; language skills were not measured in this site (Olds et al. 2004b). At age six, the total sample of participating children in the Memphis program obtained higher scores on tests of intellectual functioning and language skills than those in the comparison group and were rated by parents and by teachers as having fewer behaviour problems. In addition, the children of mothers who were considered to have mental health problems obtained higher scores on tests of arithmetic skills and verbal expressive ability than did the remaining group of participant children (Olds et al. 2004a). At age 15, there were no differences between the total Elmira sample and the comparison group children in teacher reports of problem behaviour, incidence of suspensions from school or reports related to delinquency. However, the subgroup of children of poor, unmarried mothers had lower rates of arrests, con- victions and violations of parole than children in the comparison group (Olds et al. 1999).

The Canadian HIPPY program is based on the American HIPPY program and uses the same approach and most of the same materials. The American program began operating in 1984 and cur- rently has 167 sites nationally (Gomby 2005). Nevertheless, only two evaluations of its effectiveness using a comparison group have been reported in peer-reviewed journals and these present mixed results. The first evaluation was a quasi-experimental study involving 515 children whose parents had com- pleted two full years of HIPPY and 516 demographically similar children selected at the time of follow- up, some of whom had received another type of pre- school experience while others had no preschool experience. In grade three and again in grade six, the HIPPY children performed better on reading and lan- guage achievement tests than children in either com- parison group, although the difference was very modest. There was no significant difference between the groups in achievement in mathematics, level of grade repetition or placement in special education classes (Bradley and Gilkey 2002).

In the second study, the effect of HIPPY on chil- dren’s cognitive development, school achievement and classroom adaptation was assessed at the end of kindergarten and again at the end of first grade for two sequential cohorts of children whose parents had been enrolled in HIPPY for two years in each of two different sites. One site involved a randomized trial for both cohorts, the “gold standard” for evaluation, while the other site used a quasi-experimental design (Baker, Piotrkowski and Brooks-Gunn 1999). No clear pattern of results emerged. Children in the first cohort bene- fited on different measures and/or at different times across the two sites. There was no difference between HIPPY and comparison children in the second cohort in one site, while at the other site the comparison children were rated higher than the HIPPY children on school-readiness and school achievement at the end of kindergarten, although there were no other differ- ences at either point. The researchers report that the inconsistent results could not be explained by any dif- ferences in the characteristics of the cohorts at either site or in the way that the HIPPY program was imple- mented between cohorts.

Parents as Teachers (PAT) is another home-visiting program with similar goals to HIPPY. Like HIPPY, it employs paraprofessionals and uses a relatively pre- scriptive curriculum and the provision of materials for the parent to use at home with the child. It was originally designed for use with middle-class families, and little research has examined its effectiveness for disadvantaged populations. Two randomized trials with low-income families in California both found small and inconsistent effects on parenting knowl- edge and behaviour and minimal impact on children’s cognitive, communication, social and self-help skills for the participating group as a whole. However, fur- ther analyses indicated that children living in Spanish-speaking Latino homes made small but sta- tistically significant gains in receptive vocabulary and social skills (Wagner and Clayton 1999).

Early Head Start, a federally funded program that began in 1995, serves children from birth to age three and their families. It permits its 700 sites to provide a primarily home-visiting program focusing on the par- ent, supplemented by two parent-child group activi- ties a month, a centre-based program for children, or a “mixed-approach program” in which home-visiting, and/or centre-based developmental programming is provided to different families or in combination to families either simultaneously or at different times. Seventeen Early Head Start sites across the country recruited twice as many eligible families as they could serve and randomly assigned them to partici- pate in the program (1,513 families) or to be part of the comparison group (1,488 families). At age three, there was no significant impact on children’s cogni- tive or language development in programs using the primarily home-visiting approach (Love et al. 2005).

Discussion

Several commentators have concluded that generally the use of primarily parent-focused initiatives as a sole strategy has minimal or no effect on vulnerable children’s development (Barnett 2002; Farran 2000; Gomby 2005; Gormley 2006). Two hypotheses have been suggested to account for these disappointing results. The first is that interventions concentrating on parenting education fail to address the contextual fac- tors influencing parenting style. They do nothing to address the immediate problems of inadequate income levels, substandard housing, and high rates of unem- ployment and are unlikely to address the high inci- dence of parental depression among parents living in poverty (Farran 2000). Yet we know that parental depression saps psychological energy and is associat- ed with lack of engagement in educational activities with children and a hostile rather than supportive par- enting style (Brooks-Gunn, Duncan and Britto 1999).

The relatively low levels of intensity of many parent- focused initiatives and/or actual exposure to the inter- vention is the second explanation for the disappointing results obtained with this approach. The effect of targeted interventions on child development is closely related to the intensity of the intervention (Barnett 2002; Gomby 2005; Gormley 2006; Ramey et al. 1995). Some parent-focused programs, such as HIPPY, provide their services for only a short time. Sustaining parent involvement also is an issue; it is essential for success because parents, not staff, provide the programming for the children. Obtaining sustained involvement from parents living in poverty and facing the various challenges associated with their situation has proven to be very difficult. In the United States, the rates of attrition prior to the intended length of the program are typically between 45 and 60 percent (Gomby 2005; Wagner and Clayton 1999). Even when parents remain in the program for its full duration, research indicates that many of them miss in-home visits, fail to attend out-of-home parent meetings and/or or fail to work with their child using the program materials between home visits (Gomby 2005; Wagner and Clayton 1999).

The educational background of the person provid- ing the intervention also may influence its success. Home-visiting programs targeting very low socio- economic neighbourhoods need visitors who can establish and maintain good rapport with parents and have the knowledge and skills to respond to family crises as they arise (Gomby 2005). These are not min- imal skills, yet many parent-focused programs rely on paraprofessionals. Overall, the research indicates that programs employing paraprofessionals tend to be least successful (Gomby 2005). A study that com- pared the effectiveness of the Nurse-Family Partnership program when delivered by nurses with a BSN degree and by paraprofessionals with no college education found that at age four the children whose families had been visited by a nurse had better lan- guage, cognitive and self-regulation skills than those whose parents had been visited by a paraprofessional (Olds et al. 2004b). These results do not necessarily imply that nurses make the best home visitors, but they do suggest the value of post-secondary educa- tion in a discipline related to the provision of human services. In summary, the benefits to children’s devel- opment obtained through parent-focused initiatives are inconsistent and usually negligible. The inconsis- tency may reflect differences in the planned intensity of the initiative, the actual amount of intervention received by the parent and/or the educational level of the person delivering the program.

Child-focused initiatives

Child-focused initiatives are based on the premise that the children’s home environment is unable to ade- quately support their optimal development and that direct work with the children is required in order to make a real difference in their developmental trajecto- ry. The five Canadian child-focused initiatives are the federal government’s Aboriginal Head Start Program and targeted prekindergarten in Alberta, Manitoba, Quebec and Saskatchewan. Table 2 provides basic information on each of these initiatives.

Table 2 Descriptions of Targeted Child Focused Initiatives

As illustrated in the table, all five initiatives provide part-day, out-of-home programs. Two initiatives, Aboriginal Head Start and ABC Head Start in Alberta, also provide parents with assistance accessing other community resources. In addition, ABC Head Start provides weekly parenting education sessions and vision, hearing, speech and general development screening for children. To the extent that they are available in the community, Aboriginal Head Start programs hire staff who have a college early childhood education credential. The children’s group programs in the other initiatives are staffed by certified teachers.

Aboriginal Head Start

The Aboriginal Head Start Program consists of Aboriginal Head Start in Urban and Northern Communities, established in 1995, and Aboriginal Head Start on Reserve, established in 1998. Both are completely funded by the federal government through financial transfers to Aboriginal nonprofit community groups or bands/First Nations that are responsible for designing and operating the serv- ice. Aboriginal Head Start in Urban and Northern Communities (AHSUNC) targets First Nations, Inuit and Métis children up to age six and their families living off-reserve, with a primary emphasis on children aged three to five; Aboriginal Head Start on Reserve (AHSOR) targets children of the same ages living on-reserve. In 2005-06, the total gov- ernment expenditure for AHSUNC was $31,214,712 for a total of 4,500 children served across 131 sites; the total for AHSOR was $50,165,212, with 9,101 children served in 354 sites (Government of Canada 2007b).

There are Head Start sites in every territory and province. The federal government requires that all sites provide the following components: (1) promo- tion of school-readiness through a structured part- day preschool program for children age three to five during the school year; (2) promotion of Aboriginal culture and language in the preschool program; (3) promotion of child and family health and nutrition; (4) assistance to families in accessing other communi- ty resources; and (5) opportunities for parent partici- pation on parent councils or other governing bodies and in classroom activities (Public Health Agency of Canada 2004). There is no required curriculum or cur- riculum approach and each site determines how best to provide the required components in its community context. To the extent that they are available, teach- ing staff in the preschool program are required to have the same level of training as required for child care centres in the jurisdiction in which the Head Start site operates (Government of Canada 2005b). In most cases, the lead teacher working with a group of children has probably a one-year college early child- hood education credential (Lynne Robertson, Public Health Agency of Canada, personal communication, February 19, 2007).

Prekindergarten

Targeted prekindergarten programs are operating in four provinces: Alberta, Manitoba, Quebec and Saskatchewan. A nontargeted four-year-old junior kindergarten program is also provided in Ontario. Alberta’s prekindergarten was initiated and developed by nonprofit organizations that operate Head Start- like group programs targeting children aged three and a half to five living in low-income neighbour- hoods and also provide some parent supports. These organizations supplement provincial grants for the provision of kindergarten, which can be used for chil- dren younger than age five under certain circum- stances, with funds from other sources such as the federal government, donations and their own fundraising activities. The largest, ABC Head Start in Edmonton, has 12 sites and operates 19 half-day classes four days a week during the school year. These classes have 16 to 18 children each and are staffed by a certified teacher and two assistants. Most children attend for a single year only and then trans- fer to kindergarten. ABC Head Start also provides vision, hearing, speech and general development assessments through partnerships with other commu- nity agencies; weekly parenting education groups, each with its own social worker; parent counselling and assistance in accessing other resources; and in-home visits to families whose children have special needs to assist parents in implementing educational activities to reinforce the classroom curriculum. In 2005-06, ABC Head Start served 304 children and their families and had a budget of $3,062,317, of which $2,580,942 came from the province (ABC Head Start 2006a, b; Kathy Lenihan, Executive Director, ABC Head Start, personal communication, February 6, 2007).

Two Manitoba school divisions, one in Winnipeg and the other in a northern part of the province with a high proportion of families who live on-reserve, have established targeted nursery (prekindergarten) programs for children age four. These programs do not receive provincial funding; the Winnipeg program is financed through an education levy on property taxes while the other program has been able to obtain federal funding (Joanna Blais, Manitoba Education and Youth, personal communication, February 12, 2007). In September 2005, the Winnipeg program was operating part-day classes in 58 schools and had an enrolment of 1,785 children. The cost to operate the program is expected to be approximately $4.1 million in the 2006-07 school year (Winnipeg School Division 2006). Similar infor- mation is not available for the program operating in the northern part of the province.

In Quebec, prématernelle for four-year-olds provides a part-day program operated by the local school board for four or five days per week over 36 weeks for chil- dren living in disadvantaged neighbourhoods. In 2006- 07, the program served 4,881 children (Dannie Giguère, Quebec, Ministère de l’Éducation, du Loisir et du Sport, personal communication, November 29, 2007). The esti- mated cost per child was $2,146 (Quebec 2007, 9). No new prématernelles for four-year-olds have been opened since 1997, but the existing ones continue to be funded (Friendly et al. in press). This “freeze” reflects two important government initiatives. The first initia- tive was the 1996 reform of the education system, which included a provision for four-year-old children living in disadvantaged neighbourhoods without a pré- maternelle to attend a regulated child care program without charge for 23.5 hours a week, with the express purpose of enabling them to participate in an educa- tional experience (Quebec, Ministère de l’Éducation 1996). The second important initiative was the adop- tion of a specific and mandatory educational program for use in all regulated child care settings, Jouer, c’est magique, which was inspired by the American High/Scope Education Approach, along with the rapid expansion of regulated child care spaces (Tougas 2002). Currently, children deemed vulnerable to developmental problems because they are living in a very low-income or refugee family or have been identified by a social service can attend a regulated child care program for five half-days a week without charge (Government of Quebec 2007).

Saskatchewan’s targeted prekindergarten program was started by the provincial government in 1993 as a pilot project in two communities. In 2006-07, there will be 119 provincially funded programs with an anticipated enrolment of approximately 1,900 chil- dren. School divisions decide if and where they want to operate targeted prekindergarten and receive $48,346 for each program regardless of the number of children enrolled. Many divisions augment the provincial grant with funds from other sources. There is also some provincial funding available in the first year of operation for start-up expenses. Classes have a maximum enrolment of 16 children in a mixed age grouping of three- and four-year-olds and operate half-days, five days a week during the school year. The province has produced a set of program guide- lines that emphasizes the use of direct concrete expe- riences and play to facilitate children’s development. Staffing consists of a certified teacher and a teaching assistant who often has a college certificate in early childhood education (ECE). Parents are encouraged to work in the classroom and/or sit on a parent council, and some programs also provide home visiting (Government of Saskatchewan 2006; Kathy Abernethy, Saskatchewan Learning, personal commu- nication, February 22, 2007).

Evaluations of Canadian child-focused programs

The Canadian child-focused early intervention initia- tives with child outcome evaluations are Aboriginal Head Start in Urban and Northern Communities, which has been evaluated twice, the ABC Head Start program in Edmonton, and the targeted prekinder- garten operated by Regina Public School District No. 4. Table 3 provides an overview of their evaluation approach, outcome measures and outcomes.

Aboriginal Head Start in Urban and Northern Communities (AHSUNC)

The National AHSUNC Evaluation, the report of which is yet to be released, involved nine sites across the country. Over a two-year period in the program, all the children apparently made gains in their overall physical health and social skills and 38 percent also exhibited enhanced literacy and numeracy skills (Lynne Robertson, Public Health Agency of Canada, personal communication, February 19, 2007). These results appear promising but unfortunately the evalu- ation lacked a comparison group of similar Aboriginal children who did not attend Head Start. This makes it impossible to isolate the effect of the program from other factors such as child maturation.

The Western Arctic Aboriginal Head Start Council (WAAHSC) is a network of eight AHSUNC programs in the Northwest Territories. One program has con- ducted two evaluation studies. The first compared the school-readiness of 31 AHSUNC children at entry into kindergarten and again in first grade with a compari- son group of Aboriginal children from the same com- munity, matched on socio-economic status, family structure and home language, who had not attended the program. On both occasions, the AHS children obtained significantly higher scores on school-readi- ness skills as measured by the Brigance Preschool Screen and Kindergarten Screen (Brigance 1998). In the second study, the achievement level in reading, spelling and mathematics, as measured by the WIAT- II (Wechsler Individual Achievement Test II) (Wechsler 2001), was assessed for 50 AHS children and a com- parison group when both groups of children were eight to ten years old. The comparison group in this study also consisted of Aboriginal children who had been matched on socio-economic status, family struc- ture and home language with the AHS children. Again, the AHS children obtained significantly higher achievement scores (WAAHSC 2007; Jennifer Chalmers, Evaluator, personal communication, February 20, 2007).

Prekindergarten

ABC Head Start is the largest in a network of prekindergarten programs in the Edmonton area, all of which evaluate the effectiveness of their children’s programs through pre- and post-tests using a stan- dard tool to measure children’s social/emotional, cog- nitive, motor and language skills and their general knowledge (Kathy Lenihan, Executive Director, ABC Head Start, personal communication, February 6, 2007). ABC reports that 265 children assessed just before their entry into kindergarten showed a 25 per- cent increase overall in their abilities in all four developmental areas (ABC Head Start 2006a).

Table 3 Research Approach Outcome Measures and Outcomes Targeted Child Focused Initiatives2

Unfortunately, the lack of a comparison group makes it impossible to isolate the effect of the program from other influential factors such as maturation.

Regina Public School District No. 4 recently com- pleted a longitudinal study that started in kindergarten with 48 children who had attended prekindergarten and 43 comparison children selected by their kinder- garten teachers as the best match possible in the par- ticular classroom to a prekindergarten graduate in terms of cultural and family background. In some cases, teachers determined that a match was not possi- ble, resulting in slightly fewer comparison children. In October, a slightly higher proportion of the prekinder- garten children was rated by the kindergarten teachers as having highly developed communication and/or social/emotional skills and a slightly lower proportion as having poor cognitive skills; there was little difference in ranking between the two groups on literacy skills. None of the between-group differences was statistically significant (Krentz, Mensch and Warkentin 2004a).

In first grade, 43 of the prekindergarten children and 23 of the comparison children were assessed at the end of the school year using the School Social Behaviour Scales (Merrell 2002) and on mastery of the first-grade curriculum using tools created by the Regina School Division. A higher proportion of the prekindergarten children obtained high or average scores on social competence, both groups obtained very similar overall average scores on antisocial behaviour and mastery of mathematics but most of the children reading below grade level were from the prekindergarten group. Again, there were no statistically significant differences between the groups (Krentz, Mensch and Warkentin 2004b). By the end of the second grade, only 15 com- parison children were available for testing. At that time, a slightly higher proportion of the comparison children obtained high or average scores on social competence while a slightly higher number of prekindergarten children were rated as disruptive. The comparison children obtained slightly higher average mathematics scores and as a group exhibited higher levels of reading skills (Krentz, Mensch and Warkentin 2005). By grade three, only five compari- son children were available (Krentz, Mensch and Warkentin 2006).

No clear pattern emerges from this study concerning significant between-group differences. However, a degree of caution must be exercised in interpreting these findings. First, the extent to which the two groups of children were similar to each other on salient character- istics such as family socio-economic status is unknown, as is whether some or all of the comparison children had received some sort of group preschool experience. Second, sample sizes, especially in grade two, are small — a factor that decreases the likelihood that statistical analyses will detect even moderately sized effects.

Evaluations of comparable American child- focused programs

Head Start

Head Start targets three- and four-year-olds living in poverty and is required by the federal government to provide: (1) a centre-based structured program for the children; (2) child health screening and referral; (3) hot meals that provide at least a third of children’s daily nutritional needs; and (4) social and mental health services for both the child and the family. The children’s group program may be offered on a part- or full-day basis and most children attend for two years.

Head Start has been extensively evaluated since its inception; unfortunately, however, many of the early studies lacked a comparison group, as did the more recent federally funded Family and Child Experiences (FACES) study, which compared Head Start children with national norms for all children (Zill et al. 2003). As a result, the effects of the program cannot be iso- lated from other factors that might influence chil- dren’s development. The early evaluations also tended to use small samples from a specific site or communi- ty, thereby raising the question of whether they were representative of Head Start sites in general.

More recent evaluations have used larger, samples from various sites across the country. Two such studies have compared siblings in an attempt to address the issue of differences in parents who do or do not participate in voluntary programs; this strategy is possible because Head Start is consistently over- subscribed. One study used data from the National Longitudinal Survey of Youth to compare 927 Head Start graduates with siblings who had not attended the program and found that participation closed about a third of the gap in vocabulary test scores at age five. The benefit of Head Start faded out for African American children after three or four years but was maintained by the other Head Start gradu- ates and was associated with a 47 percent decrease in likelihood of repeating a grade in either elemen- tary or high school. The positive effect of Head Start for all children exceeded that of enrolment in any other preschool program (Currie and Thomas 1995). Another study compared 489 adults from different parts of the country who reported having attended Head Start with siblings who reported that they had not. It found that White children who had attended Head Start were significantly more likely to complete high school and attend college but this did not hold true for African American graduates. They were, however, significantly less likely to have been booked or charged with a crime than siblings with- out Head Start experience (Garces, Thomas and Currie 2002).

The apparent fade-out of benefits for African American children found in the two above-mentioned studies may reflect the children’s elementary school experience, not the effectiveness of Head Start. Two studies examined the quality of the elementary school attended by children from similar low-income fami- lies who had or had not participated in Head Start. Both found that the elementary schools attended by the Head Start graduates were of much poorer quality in terms of variables such as safety, teacher-student relations and the academic climate than those attend- ed by other children from similar low-income back- grounds (Currie and Thomas 2000; Lee and Loeb 1995). Currie and Thomas found that the difference in school quality between Head Start and non-Head Start children was greater for African American chil- dren. A subsequent elementary school experience that fails to support children’s development adequately may erode the benefits of Head Start.

Sibling studies do not take into account the possi- bility that family poverty may be deeper when one sibling is a preschooler than when another is, with the result that the two siblings may have different home environments and experiences. Random assignment of eligible applicant children to Head Start or to a com- parison group increases the credibility of the findings by increasing the probability that both groups are comparable at the beginning of the study. This approach has been used in two studies. In the first study, an identification number was randomly assigned to each eligible child applicant and the chil- dren were then randomly divided into 80 Head Start and the 41 comparison group children. Assessments of all the children and family background data collected prior to the beginning of the study found no statisti- cally significant differences between the groups in parental education levels, family structure, or child receptive vocabulary or early literacy skills. Child assessment after nine months of participation in Head Start found a significantly greater improvement among the Head Start children in receptive vocabu- lary and ability to recognize and make the sounds related to letters but not print awareness or social skills (Abbott-Shim, Lambert and McCarty 2003).

In 2002, the US federal government initiated the longitudinal Head Start Impact Study. In so doing, it recognized the need for a large nationwide sample, randomized assignment of equally eligible children from the same neighbourhood to participate in Head Start or not, and follow-up in elementary school. The evaluation involves 2,449 children aged three and 2,108 aged four drawn from 23 communities (Office of Planning, Research and Evaluation, US Department of Health and Human Services 2007). The results of the first evaluation, based on data obtained nine months after the children’s enrolment, show signifi- cantly though modestly higher scores for Head Start participants on letter-word identification, prereading tests, prewriting tests and vocabulary tests and reduced behaviour problems among children who entered at age four. No significant between-group differences were found for mathematics skills (Puma et al. 2005).

Overall, the American findings on Head Start from the more recent, larger and methodologically more rigorous studies are encouraging and data on the longitudinal federal evaluation now underway will eventually enable a firm conclusion. Meanwhile, several researchers agree that Head Start has the potential to benefit vulnerable chil- dren’s development and enhance their school-readi- ness (Currie 2005; Garces, Thomas and Currie 2002; Gormley 2006; Hustedt and Barnett 2005; Zigler and Styfco 1996).

Prekindergarten

In 2005-06, 39 states funded prekindergarten programs for a total of 950,000 children, an increase of 40 per- cent in the number of four-year-olds being served since 2001-02, which meant that prekindergarten now served a larger number of children than Head Start. Most state- funded prekindergartens target children living in fami- lies with a very low income although four states now have universal prekindergarten (Barnett et al. 2006).

The National Institute for Early Education Research at Rutgers University has conducted evaluations of state- operated targeted prekindergarten in four states. All four studies employed a regression-discontinuity design5 to obtain a comparison group. This design addresses the two problems of possible selection bias and increased skills due to maturation alone. All the studies also used random selection procedures to obtain the group of children who had participated in prekindergarten and the comparison children without this experience. The number of children with prekindergarten experience in the four studies varied from 1,170 to 384 (Hustedt et al. 2007; Lamy, Barnett and Jung 2005a, b, c).

Statistically significant outcome differences between the children with and without prekindergarten experi- ence were found in all four evaluations. The increase in vocabulary over the year attributable to attending the preschool varied from 24 to 42 percent and the increase in prereading skills from 42 to 64 percent. The three studies that assessed premathematical skills reported an increase in skills of between 24 and 64 percent. None of the four studies found significant between-group differences in children’s ability to recognize, sound out or blend the sounds associated with letters. It is impor- tant to note that the comparison groups included chil- dren who had attended other types of group early childhood programs such as Head Start and child care. Thus the results isolate the effects of a specific type of intervention rather than the influence of any group program prior to entering kindergarten.

A study using data from the US National Early Childhood Longitudinal Study on a nationally repre- sentative sample of 10,224 children and controlling for a wide variety of family and neighbourhood character- istics compared reading and premathematical skills at entry into kindergarten between children who had one of the following types of nonparental experience: (1) prekindergarten; (2) Head Start; (3) another group pre- school experience such as child care or nursery school; and (4) care from a relative or nanny. Prekindergarten participation was associated with higher scores on measures of reading and mathematical skills just after entry into kindergarten than were obtained for the programs categorized as other group preschool expe- rience, but other group preschool experience pro- grams, in turn, yielded greater benefits than did Head Start or relative or nanny care. Kindergarten teachers, however, assessed children with prekindergarten or other group preschool experience as presenting more classroom behaviour problems such as poor self-reg- ulation. Behavioural concerns were not identified for the subsample of prekindergarten children who had attended a school-operated program. The researchers suggest that behaviour problems “are not a necessary consequence of pre-kindergarten” (Magnuson, Ruhm and Waldfogel 2007, 50). Instead, they may reflect lack of exposure to school classroom behavioural expectations among children whose prekindergarten program was not operated by a school. The pre- academic benefits from prekindergarten were greater for children whose parents had low educational levels and/or lived in poverty than for the full sample. The positive associations between prekindergarten and academic achievement scores had largely dissipated for the sample as a whole by the spring of first grade, but the fade-out was less for disadvantaged children (Magnuson, Ruhm and Waldfogel 2007).

The findings from the prekindergarten studies dis- cussed above are very encouraging except for the indication in the Magnuson, Ruhm and Waldfogel study that achievement score differences between children who had and had not participated in public prekindergarten fade quickly. A similar pattern of fade-out related to achievement test scores has been observed in the Chicago Child-Parent Centers pro- gram, which includes a prekindergarten component and is discussed in the next subsection. Nevertheless, longitudinal follow-up found that children who had participated in the program still had significantly higher rates of high school completion at age 20 than individuals in the comparison group and lower rates of grade repetition and use of remedial education (Reynolds et al. 2002).

Discussion

The more recent, more rigorous evaluations of American Head Start and the evaluations of the Canadian Western Arctic Aboriginal Head Start and American prekindergarten programs have consistent- ly found that centre-based preschool group programs enhance vulnerable children’s school-readiness at entry into kindergarten. The evaluation of the Regina preschool program is an exception. The findings are in the positive direction, but not significant. The lack of significance may reflect the small sample sizes and/or the possibility that the participant and com- parison groups were not comparable, a possibility that cannot be confirmed due to the lack of socio- demographic information on the two groups.

There is some indication that the American Head Start program may have less of an effect on the development of its participants in relation to that of comparison children than its prekindergarten coun- terparts. This seems counterintuitive given that Head Start, unlike prekindergarten, provides a variety of health and social services for participating children that would not be available to the majority of com- parison children. Several researchers have suggested that Head Start has lower overall quality than the prekindergarten programs operated by school boards (Hustedt and Barnett 2005; National Institute for Early Education Research [NIEER] 2003). Research shows that preschool programs of any type produce the strongest effects on child development when teachers are well qualified, and that those with a BA in early childhood education are the most effective (Ackerman and Barnett 2006; NIEER 2003; Whitebook 2003). Only about one-third of all lead teachers in the American Head Start program have this level of post-secondary education and in some states less than 15 percent of them do (NIEER 2003). On average, American Head Start teachers earn about half the salary of a kindergarten teacher in the same state, a factor likely to contribute to high teacher turnover rates in that program (NIEER 2003). In most states, teachers in prekindergarten programs receive the same compensation package as kinder- garten teachers in the same jurisdiction (Barnett, Lamy and Jung 2005; Hustedt et al. 2007). The cru- cial role in enhancing children’s development played by contributors to quality such as staff training and the number of children for whom an adult is respon- sible is discussed further in the following section.

Two-generation initiatives

Two-generation initiatives are based on the premise that changing vulnerable children’s developmental trajectories requires addressing the multiple issues contributing to their vulnerability and that this neces- sitates providing services to both the child and the parents. They typically have three components: a group program for children; a parenting education program intended to enhance parenting style and encourage parental involvement with the child in educational activities; and adult programs, such as lit- eracy and job skill training, intended to improve par- ents’ life chances. The degree of focus on each component varies across initiatives. Some focus more of their attention on enhancing child development either directly through group programming or indi- rectly though parent education than they do on adult services such as debt counselling. Others put much of their energy into adult programs and/or broad com- munity development and much less into children’s programming. Canada has three initiatives that can be considered two-generation programs: the federal gov- ernment’s Community Action Program for Children (CAPC); the Ontario government’s Better Beginnings, Better Futures Program; and the Parenting and Family Literacy Centres operated by the Toronto District School Board. Table 4 provides an overview of the approach in each of these initiatives.

Table 4 Descriptions of Targeted Two Generation Initiatives

As illustrated in the table, all three initiatives pro- vide group programs for children, parenting educa- tion, adult education and information about other community resources. However, the extent to which a site focuses on a particular type of service provision varies: the Toronto Parenting and Family Literacy Centres have the most intense focus on children’s pro- gramming and CAPC probably the least.

The Community Action Program for Children (CAPC)

The federal government’s CAPC program provides long-term funding to community-based groups and coalitions to develop and implement programs for chil- dren from birth to age six and their families living in situations that may hinder children’s development. It targets children living in low-income and/or teenage- parent families, children experiencing developmental delays and/or with social/behavioural problems, and neglected and abused children.

In a typical month in 2005-06, approximately 440 projects served 67,884 different participants. The CAPC expenditure through Health Canada in 2005-06 was $60,867,980 (Government of Canada 2007b). Historically, CAPC projects also have obtained funds from other sources such as other departments in the federal govern- ment; provincial, territorial, regional and municipal gov- ernments; donations; and their own fundraising. In 2004-05, these other sources contributed $22,946,537 (Public Health Agency of Canada 2006).

Depending on the project, program components include one or more of: (1) parenting education through home visiting or group courses; (2) group programs for children and their parent or caregiver; (3) group programs for children only; (4) individual programs for children with developmental delays; (5) adult literacy and job skills training; (6) information about community resources; and (7) community out- reach and development. CAPC projects are encour- aged to deliver the mix of components that best addresses the needs and preferences of their commu- nity. As a result, there is no single CAPC model.

Ontario’s Better Beginnings, Better Futures program

Better Beginnings, Better Futures, which began oper- ation in 1994, is a 25-year longitudinal demonstra- tion project targeting children living in communities whose characteristics may put them at risk for devel- opmental delay. There are eight sites, each in a differ- ent community; five focus on children from birth to age four and the other three on children age four to eight. In 2005-06, the Ministry of Children and Youth Services’ (MCYS) expenditures for all eight sites were $4.1 million, with an additional $800,000 provided by the Ministry of Education (Helen Hodgson, MCYS, personal communication, April 11, 2007). The five sites serving children up to age four served a total of 1,952 families and 2,667 children (Helen Hodgson, MCYS, personal communication, April 11, 2007).

Rather than requiring specific program compo- nents, the Ontario government encourages each site to develop the types and mix of services best suited to the community. The sites serving children from birth to age five provide an example of the resultant variation; one site puts much of its emphasis on direct programming with children while another has a strong emphasis on broad community development. Home visiting is provided in all five sites targeting children under age four but there is between-site variation in the background and training of the home visitors, the frequency of home visits and the age up to which visits were made (Doherty 2001).

Toronto Parenting and Family Literacy Centres

The Toronto District School Board operates 54 Parenting and Family Literacy Centres, each of which is located in a high-density, low-income, culturally diverse neighbourhood and is open to any family with a child under age five. All centres operate five days a week during the school year from 9 a.m. to 2:30 p.m. On average, families attend two or three times a week, often starting when the child is an infant, and stay for two to three hours each time. The budget in fiscal 2005-06 was $2.8 million and over the year just over 10,500 children and close to 9,000 adults attended one of the centres (Ruth Sischy, Toronto District School Board, personal communica- tion, March 6, 2007).

Each centre provides the same mix of services: (1) a group program for children, with parents/caregivers required to participate along with their child; (2) par- enting education, primarily through modelling by staff during the group program, supplemented by parenting workshops when specifically requested by parents; (3) information about other community resources; (4) a lending library with books in the home languages used in the neighbourhood; and (5) adult literacy and numeracy courses when requested by participating adults. The children’s group activities are provided by staff with at least a two-year post- secondary credential in early childhood education, most of whom also have prior experience working with families (Ruth Sischy, Toronto District School Board, personal communication, March 6, 2007).

Evaluations of Canadian two-generation initiatives

Evaluations have been conducted on all three Canadian two-generation programs. Table 5 provides a summary of the evaluation approach and sample, outcome measures and outcomes.

Better Beginnings, Better Futures

Individual sites in Ontario’s Better Beginnings, Better Futures program vary in the mixture of services pro- vided and the emphasis placed on each service within a given site. The discussion in this report focuses on the findings from the evaluation of the five sites serv- ing children under age four and their families. The researchers used two quasi-experimental designs. The first, a “baseline-focal design,” involved a comparison between baseline data collected on children age 48 months and their families in each of the five sites before the local program became operational and then five years after the program started on a different group of four-year-olds and their families in the same neighbourhood. The second approach, a “longitudinal comparison site design,” involved recruiting a group of infants and their families in each of the five target sites and in three comparison neighbourhoods without a Better Beginnings, Better Futures program plus a fourth other community. Data on the children and their families were collected repeatedly between 1994, when the children were three months of age, and 1998 to determine whether there were any changes result- ing from living in a Better Beginnings, Better Futures neighbourhood; data were also collected from chil- dren, families and teachers in the other four sites without the program (Peters et al. 2000).

Table 5 Research Approach Sample Measures and Outcom es Targeted Two Generation Initiatives

The greatest benefit to children’s development occurred in two sites: Kingston and Walpole Island. In the Kingston site, the baseline focus design found a significant decrease in neighbourhood chil- dren’s behavioural problems and increases in their prosocial behaviour and school-readiness, as assessed by junior kindergarten teachers between 1993-94 and 1997-98. The longitudinal comparison site design found that Walpole Island children obtained higher scores than children in any other Better Beginnings, Better Futures site in relation to the comparison site in motor skills and expressive language. As noted earlier, each site chose its own mix of programs and focus. The Kingston site invested extensive program resources in child care, both by enriching local child care centres in the neighbourhood and also by pro- viding a large number of informal group programs for children from birth on.

The Walpole Island community had what the researchers classified as a high-quality child care facility that was attended by over 50 percent of the children participating in the research at age 48 months (Peters et al. 2000, 31). Children in the other three sites had less exposure to direct intervention through group programs as a result of those three sites’ putting more emphasis on community develop- ment and home visiting than direct programming with children. The researchers concluded that “short- term outcomes were greatest in the area of program focus, with child-focused programs affecting child outcomes, parent/family focused programs affecting parent and family outcomes” (Peters et al. 2000, 56).

CAPC

The CAPC National Evaluation involved 1,000 families from 141 different sites across Canada and a compari- son group of 1,651 NLSCY families matched on level of household income, family structure and parental educa- tion. Interviews with families who enrolled in CAPC in 1995 or 1996 and interviews with families participating in the first cycle of the NLSCY in 1994 provided base- line data. Both the CAPC and the NLSCY parents were reinterviewed two years later. The outcome measures used related to three areas: (1) child motor and social development and indication of emotional/behavioural problems; (2) parenting skills, specifically the extent of positive support and of negative engagement with the child; and (3) the extent of maternal depression and family dysfunction. At the 24-month follow-up, there were no statistically significant differences in the health and functioning of the CAPC families and the NLSCY families (Boyle and Willms 2002).

CAPC sites vary in their primary objectives, the mix of services they provide and the degree of emphasis they put on each service. The initial statisti- cal analyses did not look for associations between the types of services children had actually received and child outcome, but this has been done in a subse- quent analysis of the same data (Palacio-Quinton 2002). Programs in which children participated, either on their own or with their parents, were associated with enhanced child development and decreases in family dysfunction; programs for adults only had the least effect on children’s development. These results are similar to those of Better Beginnings, Better Futures in their finding that involving children in direct programming had the greatest effect on their development.

The Toronto Parenting and Family Literacy (PFL) Centres

Data were collected from 108 children who had par- ticipated in one of 10 different centres prior to school entry and 108 comparison children, randomly select- ed from the same schools, who had not attended a PFL centre. When the children were assessed at age four, just after entry into junior kindergarten, using the Early Development Instrument (EDI) (Janus and Offord 2007), 55 percent of the comparison children received a low score on social competence, 51 percent received a low score on language/cognitive develop- ment and 54 percent received a low score on commu- nication skills/general knowledge. In comparison, no more than 30 percent of the children who had partic- ipated in the program received a low score on any of these three scales. The between-group scores on each scale are statistically significant (Yau 2005). Seventy- one percent of the children in each of the experimen- tal and comparison groups had a home language other than English (Yau 2000).

Evaluations of comparable American two- generation initiatives
Reviews of American two-generation initiatives note that, while several have reported effects on parents, very few have demonstrated any effects on children (Barnett 2002; Farran 2000; St. Pierre, Layzer and Barnes 1998). Differences in child outcomes among these initiatives appear to be related to the extent to which the intervention provides direct services to children. This is illustrated by comparing the Chicago Child-Parent Centers and the Comprehensive Child Development Program, as is done below.

The Chicago Child-Parent Centers

The Chicago Child-Parent Centers, established in 1967, serve children aged three to nine and their families through home visiting, parenting education, helping parents to access other community resources, child health screening and remediation such as speech therapy, a half-day preschool five days a week during the school year when the child is aged three and four, a six-week summer program between the two years of prekindergarten, and a primary grade program up to and including grade three through which the children are enrolled in classes of reduced sizes a teacher’s aide in each class, extra instructional supplies and ongoing staff development (Reynolds, Miedel and Mann 2000).

A follow-up on 934 children who participated in the program and a nonrandomized matched comparison group of 504 children from similar socio- demographic areas without a center found that when the individuals were age 21 there was a 41 percent reduction in receipt of special education, a 40 percent reduction in grade retention and a 49.7 percent high- er rate of high school completion among the partici- pants (Reynolds, Miedel and Mann 2000). Although 95 percent of the participants were African American, the fade-out of developmental gains experienced by African American participants in Head Start, as noted above, did not occur in this program. The kinder- garten and probably higher-quality grade one to three classrooms provided by the Chicago Child- Parent Centers may be a factor in this difference.

The Comprehensive Child Development Program

The Comprehensive Child Development Program (CCDP) targeted low-income children from birth through age five and their families. Parents received biweekly home visits by a case manager who assessed parental needs, referred families to other services, obtained specific services for par- ents such as adult literacy education, vocational training or employment counselling and some- times provided counselling themselves. Parenting education was provided through biweekly home visits by an early childhood specialist between the child’s birth and age three; these visits focused on the parent and did not include direct work with the child. The program was mandated to ensure that all children aged three to five received devel- opmentally appropriate early childhood education. Some sites operated their own centre-based pro- grams, while others relied on community child care programs.

Overall, the average participation in group early childhood education programs was two days a week. Thus most children had relatively little exposure to direct programming. At the beginning of the study, CCDP sites recruited twice as many eligible families as they could serve and then randomly divided the families into a group that would receive the program and a comparison group. The evaluation conducted at the end of the five-year program involved 2,213 par- ticipating children and 2,197 comparison children. At that time, there were no statistically significant between-group differences in child cognitive, lan- guage, emotional and prereading or arithmetic skills or in parenting style or employment status (Goodson et al. 2000).

Early Head Start

Early Head Start sites are permitted to choose whether to provide a primarily home-visiting program, a struc- tured, centre-based child development program, or a “mixed-approach program” in which home-visiting and/or centre-based developmental programming is provided to different families or in combination to families either simultaneously or at different times. Seventeen Early Head Start sites across the country recruited twice as many eligible families as they could serve and randomly assigned them to participate in the program (1,513 families) or be part of the comparison group (1,488 families).

At age three, when the intervention ended, there were no significant differences in child outcomes between children who had attended a site using a centre-based approach and children in the comparison group, but children who had attended mixed-approach sites obtained significantly higher scores on a measure of receptive language than the comparison group. It is important to note that in practice children from a site using a centre-based approach and children from a mixed-approach site had obtained almost identical amounts of group programming; an average of 1,400 hours in mixed-approach sites and 1,391 hours in cen- tre-based approach sites. This occurred because 30 per- cent of families in mixed-approach sites received Early Head Start Center care and other families used com- munity child care services that worked with the Early Head Start program. Parents in the mixed-approach sites received an average of two to three home visits a month while those in the centre-based approach sites had a minimum of two home visits a year (Love et al. 2005). The researchers note that there were only four centre-based sites in the evaluation and that a small sample size reduces the likelihood that statistical analysis will detect even moderate effects.

This evaluation supports the importance of direct programming for children and does not provide suffi- cient evidence to indicate that the addition of home visits significantly increases an intervention’s ability to enhance vulnerable children’s development. A follow- up of the children at age five found that the Early Head Start children were significantly less likely to exhibit behaviour problems and significantly more likely to have a positive approach to learning. The greatest effects were found among five-year-olds who had experienced both Early Head Start and a formal group program such as Head Start, prekindergarten or centre- based child care at ages four and five (Administration for Children and Families 2006).

Discussion

Two-generation initiatives vary considerably in the proportion of resources they dedicate to parental or child interventions. Those associated with the greatest child developmental gains — some CAPC sites; two of the Better Beginnings, Better Futures sites; the Toronto Parenting and Family Literacy centres; and the Chicago Child-Parent Centers — provide the great- est amount of direct programming for children. The Chicago Child-Parent Centers project used hierarchi- cal linear modelling (HLM) to determine the relative impact of different components of its program and concluded that preschool participation was more important than any of the other factors (Clements, Reynolds and Hickey 2004).

The same conclusion was reached by Project CARE, a second-generation abecedarian project involving the addition of a family education compo- nent to the center-based intervention. In this project, families received an average of 2.7 home visits a month, starting in the children’s infancy, and month- ly workshops on child development continuing until the children were on average age four and a half. At this age, the children from families that had received both home visits and centre-based programming obtained significantly higher scores than children in the comparison group on standard tests of language and cognitive skills but the two-generation interven- tion was no more successful in enhancing children’s development than the original centre-based program had been (Ramey et al. 1985).

The relative effectiveness of targeted initiatives

The relative effectiveness of parent-focused, child- focused and two-generation interventions in enhanc- ing vulnerable children’s development is a key issue for governments and service providers. A “statistically significant” difference between the children who had received an intervention and their comparison group indicates that the association between program partic- ipation and child outcome did not occur by chance, but it does not enable one to compare different inter- vention approaches in terms of their relative effective- ness. This requires converting the outcome findings from different interventions into a standard measure called the “effect size.” An effect size is the difference in means for an intervention and comparison group on an outcome variable divided by the standard devi- ation. By convention, effect sizes under 0.20 are con- sidered negligible, those between 0.20 and 0.49 are considered small and those between 0.50 and 0.79 are medium (Karoly, Kilburn and Cannon 2005, 64).

Table 6 summarizes the effect sizes that have been calculated for the American interventions discussed above. The effect size for each program was calculated on the basis of the evaluations cited in this report — that is, the effect size for HIPPY is based on the evalu- ations conducted by Bradley and Gilkey (2002) and Baker, Piotrkowski and Brooks-Gunn (1999). When reviewing table 6, it is important to note that the effect sizes were all calculated on the basis of tests administered immediately after the conclusion of the program so, while informative, they do not indicate long-term effectiveness. Also, variations in the way the child outcomes were measured mean that the effect sizes may not be strictly comparable; neverthe- less, they do provide a sense of relative effectiveness.

The Comprehensive Child Development Program, unlike the other two-generation programs, had a mini- mal child-focused component. Its lack of impact rein- forces the evidence of the effect sizes associated with child-focused programs, that vulnerable children’s devel- opment is best enhanced through participation in a group program. The table also underlines the difference in level of effectiveness between Head Start and prekindergarten even though both provide structured, centre-based programs with the specific goal of enhanc- ing children’s development and preparing them for school. The effectiveness difference is hypothesized to be associated with Head Start’s chronic underfunding, its poor levels of staff training related to child development and how to promote it, and its low compensation levels which are associated with high staff turnover rates and, as a result, lack of long-term relationships between staff and children (Hustedt and Barnett 2005; National Institute for Early Education Research 2003). No calcu- lations of effect size for Canadian initiatives were found.

The cost and coverage of Canadian targeted initiatives

Table 7 (see page 28) provides the best available information on federal and provincial/territorial expenditures for targeted initiatives to promote the development of vulnerable children and the number of children/families receiving assistance. The data pertain to the fiscal year 2005/06 except for SIPPE, Quebec’s prématernelle, the two Saskatchewan pro- grams and the expenditures for CAPC from sources other than Health Canada; in each of these cases, the relevant fiscal year is indicated on the table. As the table shows, the federal and provincial/territorial governments combined contributed approximately $243 million to targeted initiatives. Additional funds were obtained from other sources such as municipal governments, donations and school boards.

Table 6 Effect Sizes Targeted Interventions1

Table 7 Expenditures on Targeted Initiatives to Assist Vulnerable Children

Only two initiatives have developed estimates of the extent to which they reach their target popula- tion. Manitoba’s Families First calculates that it serves 81 percent of eligible women who agree to participate in the program (Mariette Chartier, Healthy Child Manitoba, personal communication, September 25, 2007); Quebec’s SIPPE estimates that it serves 56 percent of its target population (Louise Therrien, Quebec, Ministère de la Santé et des Services sociaux, personal communication, September 12, 2007). Both of these initiatives are two-generation programs. Estimating the probable coverage by targeted initia- tives as a group is impossible given the data avail- able. Some programs have specific eligibility requirements over and above having a child under age six and living in the targeted low-income neigh- bourhood; others are open to all families with an age-eligible child but target only specific low-income neighbourhoods within a low-income community.

Lessons learned

This section has focused on large, community-based targeted early childhood initiatives to enhance the developmental trajectories of children deemed vul- nerable to poor development for environmental rea- sons. In so doing, it does not present the findings of small demonstration models such as the Perry Preschool Project or interventions with low-birth- weight infants or children with physical disabilities or conditions such as Down’s Syndrome or fetal alcohol syndrome (FAS).

Sufficient longitudinal, rigorous research has been accumulated to build confidence in the following conclusions:

  1. Positive effects on vulnerable children’s development are best achieved from initiatives that target children directly with structured, centre-based pro- grams (Gomby 2005; Ramey et al. 1995; St. Pierre, Layzer and Barnes 1998).

  2. Although parent/family-focused interventions may benefit parents by, for example, increasing self- confidence, targeting children’s development indi- rectly through attempts to change parenting style and/or improve parental education and/or parental employability generally has negligible effects on children’s development (Barnett 2002; Farran 2000; Gomby 2005; Gormley 2006; Ramey et al. 1995).

  3. Quality matters. The effectiveness of group program- ming for children depends upon its quality — the extent to which staff understand child development and can translate this knowledge into effective pro- gramming, are not responsible for too many children and thus can provide individualized attention, and engage children in appropriate levels of linguistic and cognitive stimulation (Ackerman and Barnett 2006; Sylva et al. 2004; Vandell 2004; Wylie et al. 2006).

  4. The characteristics of the child’s elementary school matter. In the Chicago Child-Parent Centers, children who participated in the preschool and kindergarten components and then in the classes with the reduced sizes, fewer children per teacher and extra instruc- tional supplies for grades one through three had sig- nificantly higher achievement scores on reading and mathematics and lower levels of grade retention at the end of grade five than children who received the same preschool and kindergarten programs but were not enrolled in the supported grade one to three classes. The difference could not be explained by greater school stability among the children in the supported classes (Reynolds 1994).

  5. Intensity matters. Group programs are more effective in enhancing children’s development when they are half-day, five-days a week than when they are half- day, twice a week; a full-day group program is more effective than a half-day program provided for the same number of days per week (Barnett 2002; Gomby 2005; Gormley 2006; Ramey et al. 1995).

All the initiatives discussed in this section focus or focused on children living in low socio-economic neighbourhoods. It is only recently that research has documented the fact that many children living in mid- dle- and upper-middle-income communities are vulner- able to developmental problems as a result of the parenting style they experience and/or the extent of parental involvement in educational activities with them. The implications of this finding are discussed in the following section, which looks at: (1) who Canada’s vulnerable children are; (2) the efficiency and effectiveness of universal programs; (3) the implications of the current low quality of Canada’s child care; and (4) conditions enabling high-quality early childhood education and care.

Universal Early Childhood Education and Care (ECEC): What Do We Know?

It has long been evident that children living in low socio-economic environments are more vul- nerable to health problems, behavioural disorders and poor school careers than children from more affluent families. Recognition of this reality led President Johnson to declare a war on poverty in 1964 and to create the American Head Start program in 1965 (Zigler and Styfco 1996). Over the next decades, a plethora of targeted initiatives for low- income children prior to school entry, along with their families, followed. Now, however, policy ana- lysts and governments in the United States are begin- ning to question the value of targeting and some states have moved from targeted to universal prekindergarten for four-year-olds (Barnett, Brown and Shore 2004).

In Canada, federal and most provincial/territorial government decisions continue to be dominated by the assumption that the majority of vulnerable chil- dren live in low-income families and therefore that child vulnerability is best addressed by targeted pro- grams. Recent Canadian research challenges these assumptions. “In actual numbers, the NLSCY data show that more than 70 percent of vulnerable chil- dren in Canada live in non-poor families” (McCain, Mustard and Shanker 2007, 77). Research also shows that low-income children make substantial develop- mental gains when they receive nontargeted ECEC services (Caughy, DiPietro and Strobino 1994; Gormley et al. 2005; Kohen, Hertzman and Willms 2002; Loeb et al. 2004; Magnuson, Ruhm and Waldfogel 2007).

This section explores two policy questions. First, should publicly funded programs to enhance chil- dren’s school-readiness target children living in fami- lies with the lowest incomes, as is currently the practice in Canada, or be universally available to all families wishing to use them? Second, what condi- tions are necessary to obtain the greatest benefit from ECEC programs? The following section examines the benefit/cost ratios associated with targeted and uni- versal ECEC.

Who are Canada’s vulnerable children?

The majority of children in Canada are born healthy, reach developmental milestones at the expected time and enter school with the level of physical, social, lan- guage, self-regulation and cognitive skills required to benefit from the school program (Canadian Council on Learning 2007). Nevertheless, five large Canadian stud- ies have reported that 25 to 30 percent of five-year- olds in the general population are developmentally delayed at school entry in one or more of these crucial school-readiness skills (Doherty 2007).6 Such children are referred to as vulnerable children. Research indi- cates that children who live in an Aboriginal family may be even more vulnerable than those in the general population. A study of 4,226 Aboriginal five-year-olds in British Columbia found that 40 percent of them received a score indicating a lack of school-readiness on at least one scale of the Early Development Instrument (EDI) (Jennifer Lloyd, Human Early Learning Partnership, University of British Columbia, personal communication, October 31, 2006).7

All types of health and social outcomes appear as a gradient when plotted against the social and eco- nomic status of the population studied, a phenome- non known as the socio-economic gradient (Keating and Hertzman 1999). Consistent with this general finding, the incidence of poor developmental out- comes in the general population at age five is highest among children living in poverty. The prevalence of vulnerability is lower among children living in mid- dle-income families and lowest in the most affluent families. However, there are fewer children living in poverty or in affluent families than in middle-income families, a group that in Canada represents about 75 percent of the total population (McCain, Mustard and Shanker 2007, 46). Targeting initiatives to enhance children’s development to children living in the low- est-income families excludes a large number of vul- nerable children and provides special programming to children whose school-readiness is not at risk. If our aim is to enhance the school-readiness of all children, this approach may not represent the most cost-effec- tive use of public money.

Universal approach to address vulnerability

The alternative to targeting is a universal ECEC sys- tem. There is sometimes confusion about the mean- ing of “universal” in this context. It does not mean a program in which all age-eligible children must participate. It does mean a service that is widely available and affordable for all families wishing to use it. Examples of universal ECEC systems are found in Belgium, Italy and France, where virtually 100 percent of all three- and four-year-olds participate (OECD 2006, 78). The only jurisdiction in Canada that comes anywhere close to having uni- versal ECEC is Quebec. In 2006, there were 204,114 regulated child care spaces in that province, includ- ing those in family child care settings (Quebec, Ministère de la Famille, des AiÌ‚nés et de la Condition féminine 2007), for almost 380,000 children (Institut de la statistique 2006). Low parent fees, as a result of high government subsidization, and the govern- ment’s waiver of parent fees for very low-income parents make the available spaces affordable for the majority of families.

While the initial cost is higher for a universal than for a targeted program, a universal approach is an attractive alternative for two reasons. First, a higher proportion of vulnerable children would be reached (Belfield 2006a).

Support for believing that more vulnerable chil- dren would be reached by a nontargeted program comes from three sources: Europe, Argentina and the United States. Most European countries provide uni- versal, publicly funded or highly subsidized ECEC for three- and four-year-olds. Participation by three- year-olds in these programs is virtually 100 percent in France, Italy and Belgium and over 70 percent in Germany, Denmark, Hungary, Norway and Sweden (OECD 2006, 78). Argentina’s expansion of govern- ment-funded preschool was associated with a jump in preschool enrolment from 49 to 64 percent (Berlinski, Galiani and Gertler 2006). Similarly, American states with universal government-funded prekindergarten programs report that they are used by 60 percent or more of age-eligible children (Committee for Economic Development 2006). The high rates of usage associated with affordable, widely available early childhood education programs increases their ability to reach a higher proportion of vulnerable children not living in poverty than services targeted to low-income neighbourhoods.

Second, there is clear evidence from Argentina, Britain, New Zealand, Sweden and the United States that the school-readiness of children from all socio- economic backgrounds and their subsequent academ- ic success can be enhanced by participation in early childhood education, whether it is prekindergarten or ordinary child care. In addition, as is the case for tar- geted programs, the returns to the public purse from universal ECEC services outweigh their costs (see table 10, page 35).

Two longitudinal British studies have documented the beneficial effect of participation in nontargeted ECEC, whether it was child care, preschool/nursery school or a formal regularly scheduled playgroup, for children from all socio-economic backgrounds. The sam- ple in the first study was drawn from all the children born in the United Kingdom in a single week in April 1970 (Osborn and Milbank 1987). When 6,261 children were assessed at age ten on five standard achievement tests, including 2,468 children who had not had any type of ECEC experience, there was a statistically significant difference on each test in favour of the children who had participated in ECEC. Similarly, the longitudinal Effective Provision of Preschool Education (EPPE) study of 2,793 children found that those who had participated in any type of ECEC entered school with higher levels of cogni- tive skills than those without this experience (Sammons et al. 2002). The beneficial effects of participation in ECEC on children’s reading and mathematical skills found at age five continued to be evident at age seven (Sylva et al. 2004). By age ten, the impact of the quality8 of the ECEC became more important than whether or not the child had ECEC experience. Children who had partic- ipated in high-quality ECEC continued to exhibit a sta- tistically significantly better reading ability and slightly better mathematical skills than children who had attend- ed a low-quality ECEC program or children without ECEC (Sammons et al. 2007).

The importance of high-quality services

While these studies collapsed various types of ECEC into a single entity, other research has focused specifi- cally on the effect of participation in either child care or prekindergarten. Longitudinal studies conducted in both Sweden and the United States comparing children who have or have not participated in child care report an association between such participation and more successful school careers (Andersson 1992; Broberg et al. 1997; Burchinal et al. 1995; Caughy, DiPietro and Strobino 1994). However, this is not always the case. Other studies have found no between-group differences or a negative association between participation in child care and children’s elementary school performance (Vandell and Wolfe 2000).

A review of the last 20 years of child care research concluded that whether or not child care promotes chil- dren’s development depends upon its quality — the extent to which the interactions between adults and children are warm and supportive and the children have ample opportunities to engage in activities that provide linguistic and cognitive stimulation (Vandell 2004).

The importance of the level of child care quality for children’s development has been demonstrated by large longitudinal studies in both New Zealand and the United States. The New Zealand study followed children to age 14 and found that the associations noted at age eight between child care quality and level of mathemati- cal skills and reading comprehension persisted (Wylie et al. 2006).

In an American study, children whose child care was rated in the highest third of quality levels obtained higher scores on tests of preacademic skills and language at age four-and-a-half than did children whose child care was rated in the lowest third. The difference in effect on children’s skills related to experiencing child care in either the highest or lowest third of quality was 0.39 for preacademic skills and 0.29 for language skills (National Institute for Child Health and Human Development [NICHD] Early Child Care Research Network 2002). At the end of grade three, and with factors related to elementary school experi- ence controlled, children who had participated in higher-quality child care obtained significantly higher scores on standardized tests of mathemati- cal skills, vocabulary and memory than children who had participated in poorer-quality child care (NICHD Early Child Care Research Network 2005). A significant positive association between child care quality and vocabulary score persisted even at the end of grade six (NICHD Early Child Care Research Network 2007).

Similarly, a second American study, Cost, Quality and Child Outcomes, found that the effect size of level of child care quality just prior to entry into kindergarten was 0.56 for sociability, 0.43 for cognitive skills and 0.18 for language skills. Participation in a higher- rather than lower-quality program was associated with higher scores on measures of vocabulary and mathematical skills and teacher ratings of the children’s sociability at the end of grade two (Peisner-Feinberg et al. 2001).

Attending Argentina’s universal prekindergarten for children from age three to five is associated with higher scores on achievement tests in Spanish and mathematics in grade three, with an estimated increase of 8 percent for each year of preschool attendance (Berlinski, Galiani and Gertler 2006). Two evaluations of Oklahoma’s universal prekinder- garten program both found higher levels of school- readiness just prior to kindergarten among children who had participated in the program in comparison to children without this experience (Gormley et al. 2005; Lamy, Barnett and Jung 2005d). Lamy and her colleagues report that the increase in school-readi- ness attributable to participation in prekindergarten was 88 percent for prereading skills, 44 percent for premathematical skills and 28 percent for vocabu- lary. Similar positive results have been found for West Virginia’s universal prekindergarten with the increase attributable to the program being 121 per- cent for prereading skills, 63 percent for premathe- matical skills and 30 percent for vocabulary (Lamy, Barnett and Jung 2005e).

Conditions enabling high-quality early childhood education and care

Studies exploring the characteristics of those ECEC programs that are the most successful at promoting children’s development have reported remarkably consistent findings whether the program is targeted or nontargeted and regardless of what it is called. The adults responsible for a group of children have a university degree and specialized training in early childhood education and are well compensat- ed, the group sizes and children-to-adult ratios allow for individualized attention and support, and the children receive a planned, purposeful, devel- opmentally appropriate daily program (Ackerman and Barnett 2006; Sylva et al. 2004; Vandell and Wolfe 2000; Wylie et al. 2006). Effective ECEC pro- grams also have a physical environment and mate- rials that support good programming practices, provide regular professional development for staff and engage in ongoing self-evaluation (Ackerman and Barnett 2006).

The characteristics noted above are important enablers of high-quality programming inasmuch as they establish conditions that permit and encour- age such programming. Post-secondary education and specialized training related to child develop- ment increase the probability that an adult will understand how young children develop, will respond sensitively to them and will also provide stimulating, developmentally appropriate experi- ences that promote their development. A purpose- ful, planned daily program increases the probability that children will engage in develop- mental activities rather than aimlessly wandering about. Smaller group sizes and fewer children per adult enable a greater number of individual inter- actions between adults and children and the tailor- ing of activities to the particular needs of each child. Adults who earn enough to live on and feel their compensation reflects their level of education and responsibility are more likely to stay. This results in more stable adult-child relationships and a greater sense of security for the children and gives the adults a better understanding of each child’s developmental level. Indoor and outdoor spaces that are safe, child-friendly, pleasant to be in, large enough to permit a variety of activities and accessible to people with disabilities support good programming, as does the availability of appropriate and sufficient materials.

The relative effectiveness of universal programs

Table 8 summarizes the data available on effect sizes for universal prekindergarten and for univer- sal ECEC programs of all types combined. The only data located on effect sizes for child care relate to the differences in the effect on children’s develop- ment directly associated with differences in pro- gram quality level (National Institute for Child Health and Human Development [NICHD] Early Child Care Research Network 2002, 2005; Peisner- Feinberg et al. 2001) or estimates of child outcome gains that would be achieved for specific incre- ments in program quality (NICHD Early Child Care Research Network and Duncan 2003). There may be some reluctance to publish effect sizes for child care based on combining data from all the child care programs that participated in a specific evalu- ation due to the generally poor quality of the child care programs evaluated. For example, only 14 percent of the centres in the Cost, Quality and Child Outcomes study were rated on a standard observational tool as providing a program consistent with supporting children’s development rather than just providing custodial care (Helburn 1995). Calculating effect sizes based on a sample with few high-quality programs would underestimate child care’s potential to enhance children’s development.

Benefit/Cost Ratios

Many of the benefits derived from ECEC pro- grams that enhance children’s development can be translated into dollar figures and compared with program costs to calculate a benefit/cost ratio. The standard methodology and eco- nomic equation used to perform a benefit/cost analysis, and thus determine the ratio, is described in detail by Aos and his colleagues (2004, appendix D).

Table 8 Effect Sizes Universal ECEC Various Jurisdictions1

When considering benefit/cost estimates, it is essen- tial to recognize that they are not the final word although they can provide a sense of the relative differ- ences across approaches. Benefit/cost analyses vary in the benefits and costs included in the calculation and across jurisdictions as a consequence of differences in factors such as the relative costs of personnel and facili- ties and whether the proposed program is full day or part day. For example, if a prekindergarten program yields the same economic returns in two jurisdictions but wage levels are higher in one, the benefit/cost ratio in that jurisdiction will be lower. Sometimes a benefit/cost ratio is projected beyond the age of final follow-up by esti- mating lifetime differences on the basis of experience in the general population such as the greater employability of people who have completed high school in compari- son with those who have not, and the lower probabili- ty that they will use social assistance.

Targeted programs

Benefit/cost analyses for targeted programs start by using the statistically significant findings related to different outcomes between a participant and com- parison group. In the case of the benefit/cost esti- mates reported in table 9, this information came from the specific program evaluations cited earlier. The benefits are based on a number of variables, the mix of which varies depending on what was meas- ured by the researchers. The possible variables include reduced expenditures for grade repetition and use of remedial education services; increased likelihood of high school graduation; savings to one or more of the child welfare, social assistance or child and/or adult criminal justice systems; and increases in government revenue through increased income tax. The Chicago Child-Parent Centers bene- fit/cost ratio also includes benefits not accruing to government such as the half-day reduction in the cost of child care for employed parents. Because benefit/cost ratios do not include other potential benefits that were not measured in the program eval- uation, such as stronger national economic competi- tiveness as a result of improvements in the educational attainment of the future workforce, ben- efit/cost estimates are likely to be conservative (Karoly, Kilburn and Cannon 2005).

Table 9 BenefitCost Estimates for Targeted Initiatives

The benefit/cost ratios presented in table 9 show a benefit of $1.23 to $7.14 for every dollar of pub- lic funds spent. The Chicago Child-Parent Centers yield the greatest return, with very clear benefits to the child’s school career. Although it is a two-gen- eration program, its primary emphasis is child- focused and it includes additional support for the elementary school classes used by participants up to and including grade three as well as health and nutrition services. The Arkansas prekindergarten program, which provides a lower benefit, provides solely a classroom experience. Of the three parent- focused initiatives, the greatest benefit to the pub- lic purse comes from the Nurse-Family Partnership. This benefit primarily reflects savings to the child welfare, social assistance and child and adult crim- inal systems and greater likelihood of maternal employment. The NFP program was not designed to enhance children’s school-readiness or school career and does not provide the same level of ben- efit in these areas as do child-focused programs (see table 6, page 27). No benefit/cost analysis for Head Start is included in either of the two docu- ments located that review benefit/cost ratios for several initiatives.

Universal prekindergarten

No benefit/cost analysis appears to have been done for any of the four existing American universal prekindergarten programs. However, six other states have estimated the likely benefit/cost ratios if they implemented universal prekindergarten. The method- ology used was similar to that used to determine benefit/cost ratios for targeted early childhood inter- vention but also included the need, in some states, to improve program inputs so that they are similar to those in the Chicago Child-Parent Centers — bachelor degree lead teachers with a specialization in early childhood education who receive the same compensa- tion packages as those received by other teachers in the system, a maximum class size of 20 and no more than 10 children to every adult (Belfield 2005, 2006b; Karoly and Bigelow 2005). As illustrated in table 10, a positive return on investment was estimated for each state, with benefits ranging between $2.62 to $1.18 for every dollar spent. The Committee for Economic Development (CED) has estimated the likely benefit/cost ratio for a nationwide prekindergarten program in the United States as at least $2 for each dollar of public funds. The CED report addresses the substantial initial additional expenses that would be incurred in implementing a universal prekindergarten program by suggesting various ways to fund it (2006).

Universal child care

Analyses of the benefits/costs of universal child care have been done in both Canada and Switzerland. In Canada, two economists from the University of Toronto estimate that a nationwide high-quality universal child care system for children aged two to five with parent fees geared to income would yield $2 for every dollar of public money spent. Their methodology, detailed in appendix B of their report, took into account factors such as costs not incurred for special education and grade repetition and increased income taxes as a result of increased parental workforce participation (Cleveland and Krashinsky 1998).

As indicated by table 10, the estimated cost per par- ticipant per year for universal child care is greater than for the American universal prekindergarten programs, even taking into account the exchange rate. There are two reasons for this difference. First, the child care day is longer than the standard eight-hour work day, result- ing in a need for additional staff to cover the beginning or the end of the day. Second, some children participat- ing in child care would be younger than age three.

Table 10 BenefitCost Estimates for Universal ECEC Various United States and Canada 

Desirable children per adult ratios are smaller for younger children, with the result that each adult work- ing with younger children should not be responsible for as many as an adult working with three- and four- year-olds. In spite of the higher cost per child, the esti- mated benefit compares favourably with the estimated benefits for the prekindergarten programs.

Using annual data from Statistics Canada’s Survey of Labour and Income Dynamics, a study in Quebec calculated that implementation of the five-dollar-a- day child care program, now seven dollars a day, was associated with a 21 percent increase in maternal paid employment outside the home and additional tax income that offset about 40 percent of the cost of the program (Baker, Gruber and Milligan 2006). Similar findings are reported from Switzerland where public investment of CHF 18 million in child care by the city of Zurich is estimated to be offset by at least CHF 29 million from additional tax revenue and reduced spending on social assistance (OECD 2006).

Discussion

Tables 9 and 10 demonstrate that the benefits to society of both targeted and universal large-scale ECEC initia- tives are greater than is their cost to the public purse. Such findings support the strong consensus among economists, biologists and social scientists that public investments in the early childhood period provide greater returns than at any other period in life (Lefebvre and Merrigan 2003; McCain, Mustard and Shanker 2007). In fact, a group of American economists has estimated that, “a dollar invested in early childhood [development] yields three times as much as for school- aged children and eight times as much as for adult edu- cation” (McCain, Mustard and Shanker 2007, 136). This is not surprising. Development is sequential and cumu- lative and it is during the preschool period that the foundation is laid for the individual’s lifelong physical and mental health, the child’s degree of success in school, and the adult’s life skills and employability. The human capital built through enhancing young chil- dren’s development assists in creating and maintaining a prosperous society and a prosperous society encour- ages tolerance and social stability.

Policy Implications and Discussion

Canada is facing a workforce shortage resulting from demographic factors. This reality, and the demands of new technologies and the global economy for workers with good people, litera- cy, numeracy, problem-solving and decision-making skills, means that Canada’s future prosperity depends upon ensuring that every child reaches his or her fullest potential (Dodge 2003; Fortin 2006; Lefebvre and Merrigan 2003). The need to do this is so critical that one economist has suggested that “our most pressing task as a nation…is to foster basic skills such as literacy and numeracy” (Fortin 2006, 5).

Successful acquisition of literacy and numeracy depends upon a foundation of competencies devel- oped prior to entry into kindergarten, yet 25 percent of children in Canada enter school lacking one or more of the essential foundation skills (McCain, Mustard and Shanker 2007). In 1999, Margaret McCain and Fraser Mustard raised the alarm about the extent of child vulnerability to developmental problems in Canada and urged governments to make a major effort to improve the opportunities for opti- mal early childhood development for all children. Nearly a decade later, the response of Canada’s feder- al, provincial and territorial governments to the need to promote every child’s development is woefully inadequate in both approach and degree.

Federally, provincially and territorially funded ini- tiatives to assist the development of children vulnera- ble to poor developmental outcomes are limited to programs targeting children living in low socio-eco- nomic neighbourhoods and to Aboriginal children. However, vulnerable children live in families across all income levels and in both the Aboriginal and non-Aboriginal populations. Although the incidence of vulnerability is highest among children living in poverty, the largest number of children, approximate- ly 75 percent, live in middle-income households (McCain, Mustard and Shanker 2007, 46). Restricting publicly funded early-intervention initiatives for vul- nerable children to low-income and Aboriginal fami- lies misses the majority of children who are vulnerable.

We also know that positive effects on vulnerable children’s development are best achieved with interven- tions that target children directly with structured, cen- tre-based group programs. Yet, as table 7 shows, over 50 percent of federal/provincial/territorial early interven- tion funds are directed to parent-focused and two-gen- eration initiatives. While such interventions can yield benefits to the individual parent, such as increased self- confidence and job skills, and to society by reducing the need for social assistance, they provide negligible bene- fit to children’s development (see table 5).

Finally, programs targeted on low-income families do not even reach their intended clientele. Only two initia- tives have developed estimates of the extent to which they reach their target population. Manitoba’s Families First reports that it serves 64 percent of eligible women, that 22 percent decline and that 14 percent are not served for other reasons (Mariette Chartier, Healthy Child Manitoba, personal communication, September 25, 2007). Quebec’s SIPPE estimates that it serves 56 percent of its target population (Louise Therrien, Quebec, Ministère de la Santé et des Services sociaux, personal communication, September 12, 2007). Both of these ini- tiatives are parent-focused programs that do not provide direct programming for children. In 2001, the most recent date for which information is available, 415,655 children under age six were living in poverty (Canadian Council on Social Development 2007, table A-1). In the intervening period, child poverty levels have only been reduced consistently in each year in Quebec, whereas in other jurisdictions change has been variable or poverty rates have increased (Campaign 2000 2006, 2). The par- ticipant statistics provided in table 7 indicate that approximately 23,000 children under age five across the whole of Canada received child-focused targeted group programs.

The realities of child care in Canada

Child care plays a significant role in the lives and thus the development of many Canadian children under age five. In 2005, 52 percent of children under age three and 54 percent of those between age four and five regularly received some form of nonparental care while their parents engaged in paid work or fur- thered their education; they spent, on average, 27 hours a week in that care (Bushnik 2006). Yet, 38 per- cent of children who receive regular nonparental care in Canada receive it in unregulated settings that do not have to meet even basic health and safety stan- dards (Bushnik 2006). This reflects the reality that, in 2006, there were regulated child care spaces for only 21 percent of children under age five (Friendly et al. in press). While participating in regulated child care protects health and safety, only about a third of such programs provide the type and amount of programming that promotes children’s social, language and cognitive development (Doherty et al. 2006; Drouin et al. 2004; Goelman et al. 2006; Japel, Tremblay and CoÌ‚té 2005).

As discussed in the previous section, the adults responsible for a group of children in the kinds of ECEC services proven to promote children’s development have a university degree and specialized training in early childhood education, are well compensated and provide a planned, purposeful, developmentally appropriate daily program. But in 2001, the most recent year for which data are available, the majority of people working direct- ly with children in Canadian child care had a one-year community college certificate or a two-year diploma; less than 15 percent had a bachelor’s degree or higher qualification (Beach et al. 2004, 16). Salary levels were low and remain so. On average, staff in child care cen- tres who had a degree earned $21,023 in contrast to the $47,146 earned by an elementary school teacher with the same level of education (Beach et al. 2004, 25). This large discrepancy does not encourage degree-holders to remain in the child care field. Unlike the situation in kindergarten, where all the provinces and territories require the provision of a specific curriculum or types of activities in every program, only Quebec requires all reg- ulated child care settings to use a specific curriculum.

The characteristics associated with effective ECEC are expensive but necessary. Yet outside of Manitoba and Quebec, child care programs have to rely on fees paid by parents for an average of almost 50 percent of their budget (OECD 2004, 57). However, parent fees cannot be sufficiently high to cover the true cost of providing high-quality ECEC — for example, to cover the cost of staff compensation that reflects the desir- able level of training. Raising ECEC fees to increase compensation levels is not an option since it would reduce access. Instead, as noted by the OECD (2004, 72) in its report on its evaluation of services in Canada for children under age six, public funding of ECEC must be substantially increased.

As a matter of comparison, child care in Europe is generally of high quality and generously subsidized by government. Parent fees for children under age three in countries such as Finland, Norway and Sweden range between 9 and 15 percent of the cost; fees are about 25 percent of the cost in the rest of Continental Europe. In most European countries at least part-day fully subsi- dized ECEC is provided for all three- and four-year-olds (OECD 2006, 78).

Poor-quality child care is not simply a missed devel- opmental opportunity; it is known to be detrimental to all children’s development regardless of their socio- economic background or the extent to which their parents engage in developmental activities with them (Howes 1990; Peisner-Feinberg and Burchinal 1997; Vandell and Corasaniti 1990).

Given the large number of children under age five regularly participating in child care, Canada cannot continue to treat this service as simply a safe place for children to stay while their parents work. It must be recognized as a developmental opportunity for children and an investment in Canada’s future pro- ductivity. Doing so would require higher regulatory standards, especially in the area of staff training, and much higher levels of public funding than currently provided outside of Quebec to enable the sort of com- pensation that encourages people to complete the required level of training and remain in the field.

What needs to be done?

Addressing the current high levels of lack of school- readiness in an effective fashion requires:

  1. accepting that vulnerability to poor developmental outcomes occurs across all income levels and, as a result, interventions must be universal since there is no easily observable neighbourhood marker to identify all vulnerable children;

  2. developing a nontargeted, evidence-based, cost- effective strategy; and

  3. providing the level of government funding required and other resources for the strategy to be effective.

This is a tall order but doable, although it will take time, and it is affordable.

Factors to consider

When considering the type of group ECEC program that would be the most effective, attention must be paid to: (1) the high workforce participation by women who have a child under age five; (2) the approach that would work best for the child and fam- ily; (3) the need for complementary services such as developmental screening prior to entry into kinder- garten; and (4) the provision of additional supports for those with unique needs.

The percentage of women in the paid workforce who have at least one child under age five has increased each year since 1976. By 2006, 64.3 percent of women whose youngest child was under age three engaged in paid employment, as did 69.4 percent of those whose youngest child was age three to five (Statistics Canada 2007, table 5). Most of these women worked full time and, increasingly, a grand- mother is not available to provide care. In fact, the labour force participation of women aged 55 to 64 has sharply increased since the mid-1970s from 32 to 51 percent (Marshall and Ferrao 2007, 6). Therefore, many parents must find reliable nonrelative care.

The second factor to consider is how best to meet the needs of both child and family. Young children thrive when they experience stable relationships with a minimal number of transitions from place to place during the course of a day and are engaged in a daily program of interesting activities that stimu- late their linguistic and cognitive development. Working parents’ primary needs are affordability, a convenient location and a service that operates for the hours required to cover the time the parent is out of the home. Access to government-funded part- or full-day prekindergarten programs for three- and four-year-olds would lessen the cost of alternative care for employed parents and, being located in the local school, such programs would probably be con- venient. However, parents who work full time require alternative care for longer than the tradi- tional 9 a.m. to 3 p.m. school day. Currently in Canada, the need for care beyond the traditional school hours is addressed by “wrap-around” child care provided before and after the prekindergarten program either in another room in the school or in a nearby location, with the child being transported back and forth. In this situation, the child experi- ences one set of adults providing “care” and another set providing “education” and several transitions to different physical locations within the course of the day. Other countries have adopted the philosophy that care and education are indivisible for young children and should be provided by the same group of adults in a program that blends educational and care activities throughout. Following this philoso- phy, several European countries provide universal ECEC programs operating for 10 hours a day that incorporate play-based learning activities through- out the whole day delivered by adults with three or four years of specialized post-secondary training to equip them to implement the underlying philosophy (Cameron 2004). This level of training is consistent with the pan-Canadian requirement, except in Prince Edward Island,9 that kindergarten teachers, whether working with four- or five-year-olds, have a four-year B.Ed. or a three-year BA and one year of specific teacher training (Doherty, Friendly and Beach 2003).

Early identification and remediation of visual or hearing impairments or speech impediments enhances the child’s chances for optimal development. Currently, in some parts of Canada, the first universal opportunity to participate in developmental screening after infancy occurs when the child enters kinder- garten. At that point, undetected hearing and visual impairments or mild speech impediments may already have had a negative effect on the child’s development.

Finally, the provision of universally available ECEC does not necessarily mean equal funding per child; the Canadian health system is universal but also recognizes that some citizens have unique needs that require additional services (Brownell et al. 2006). A case can be made that the principle of additional supports for those with unique needs may apply to two specific groups of young children — those living in Aboriginal families and those whose home lan- guage is not that of the local school.

British Columbia’s province-wide EDI study reports that 40 percent of Aboriginal children obtain scores indicating a lack of school-readiness on one or more scales, with the lowest scores being obtained on the language/cognitive development and general knowl- edge/communication skills scales (Jennifer Lloyd, Human Early Learning Partnership, personal commu- nication, October 30, 2006).

A Toronto study found that when a group of low- income five-year-olds, 71 percent of whom had a home language other than English, were assessed by the EDI, just over half of the children obtained scores suggesting a lack of school-readiness in one or more of the social competence, language and cognitive development, or communication skills/general knowl- edge scales, a much higher average than found in the population as a whole (Yau 2005, 2).

The adjustment to and later success in school of both these groups of children might be enhanced not only by early exposure to the language used in the local school but also by assistance in understanding and adapting to its norms and experiences that pro- mote early literacy and numeracy skills. The provi- sion of additional services might include additional government financial support to ECEC programs operating in neighbourhoods with a high proportion of families whose home language is neither English nor French or of Aboriginal families, so that the pro- grams can operate with fewer children per adult and the staff receive specific training related to address- ing each group’s unique needs.

An example of what might be done

The recent follow-up to the 1999 Early Years Study proposes a universal system of early child development and parenting centres that would be affordable and widely available to pregnant women and families with a child under age six (McCain, Mustard and Shanker 2007). Each centre would incorporate pre- and postna- tal supports, parenting education, developmental screening, child care and kindergarten and report to a single local authority that would be administratively responsible for the group of centres. Ontario’s Best Start Initiative, announced in 2004, is already moving in this direction. Three pilot projects in three very dif- ferent parts of the province are developing community hubs through formal coordination and collaboration among existing organizations so that families have a single entry point to a continuum of ECEC services and other services such as developmental screening. Currently, there are 24 such hubs (Ontario, Ministry of Children and Youth Services 2007a). Best Start’s ulti- mate goal is full integration of child care and kinder- garten into a single ECEC service that works in close partnership with the local public health unit and other community organizations to provide all their services in a single location. As noted by the province, achiev- ing this goal will require a transformation of the way in which services are planned, funded, managed and delivered. A detailed timetable including 15 measurable goals, each with its own set of tasks and target dates, has been developed to guide the transformation (Ontario, Ministry of Children and Youth Services 2007b). Meanwhile, Ontario is addressing the immedi- ate need to improve the quality of regulated child care.

The bottom line

What would it cost the public purse to do what needs to be done? Is the cost affordable? Taking into account the services that already exist and are publicly funded, McCain and her colleagues estimate that their suggest- ed model would require about $8 billion in new spend- ing (McCain, Mustard and Shanker 2007, 142). This includes improved workforce training and compensa- tion, both of which should also improve program quali- ty and thus child outcomes, and the assumption that parents would contribute to the cost through fees geared to income. As noted in tables 9 and 10, the American experience indicates that the benefit of ECEC programs to the public purse outweighs their cost. In Canada, approximately 40 percent of the cost of Quebec’s seven-dollar-a-day child care program is being recovered through additional income tax revenue from increased maternal engagement in the paid workforce (Baker, Gruber and Milligan 2006). The steady decline in child poverty associated with increased maternal employment has also meant a steady decrease in the use of social assistance in Quebec (Campaign 2000 2006). Thus Quebec is already receiving significant benefits from its ECEC expenditures and stands to recoup more of its invest- ment in the future through increased rates of high school completion and a workforce that is better able to compete successfully in the global economy.

Conclusion

Canada’s future prosperity depends upon its ability to develop the knowledge and skills of what will be a smaller workforce in the com- ing decades (Brownell et al. 2006; Dodge 2003; Fortin 2006; Lefebvre and Merrigan 2003). The research is clear: the foundation for school success and adult well-being and competency is laid down during the years prior to school entry (McCain, Mustard and Shanker 2007). The implication is also clear: Canada must do all it can to ensure the optimal development of every child. While the incidence of children vul- nerable to developmental problems is highest in the lowest socio-economic group, the largest number of children overall live in middle- and upper-middle socio-economic families and this is the group where most vulnerable children are found (McCain, Mustard and Shanker 2007). Restricting early intervention ini- tiatives to low-income neighbourhoods misses the majority of vulnerable children. Failing to address the reality that over 50 percent of children under age four regularly receive out-of-home child care, most of which provides insufficient activities to support their development, puts many children at risk for vul- nerability who might not be so otherwise. It is time to recognize that supporting the development of all chil- dren requires a system of high-quality ECEC that is available and affordable to all families wishing to use it and to act on this recognition.

The OECD suggests that 1 percent of the gross domestic product (GDP) is the minimum requirement for an ECEC system of sufficiently high quality to enhance children’s development and be affordable and available to all families who wish to use it. Currently Canada’s annual expenditures on ECEC are about 0.3 percent of its GDP ((2006, 105).10 The cost to Canada of spending 1 percent of its GDP for ECEC for children under age six would be about $10 billion (McCain, Mustard and Shanker 2007, 142). Universal ECEC provides greater benefits to society than it costs in public expenditures. Canada can afford this expen- diture; Canada cannot afford the inevitable conse- quences for its future prosperity of failing to implement a universal ECEC system.

  1. Quebec agrees with the objectives of the National Children’s Agenda. However, the Government of Quebec has decided not to participate in its develop- ment because it wishes to assume full control over programs aimed at family and children within its terri- tory (Federal-Provincial-Territorial Council of Ministers on Social Policy Renewal 1999, 1).

  2. Research consistently shows that how the parent inter- acts with the child has a strong influence on the child’s development and later school career. Warm, responsive and supportive parents who set clear behavioural limits but allow children to have a voice in what happens are associated with positive school outcomes; in contrast, children who experience highly controlling and harsh parenting are less likely to do well in school. Children’s development is also associated with the extent and type of verbal and other stimulation, such as opportunities to play with manipulative toys, available in the home. Parent-focused programs encourage parents to be warm, supportive and consistent with their children and provide them with information about how to sup- port and stimulate their child’s development.

  3. A quasi-experimental evaluation involves comparing the outcomes of two groups of people, one that partici- pates in an intervention and the other that does not, matched with each other on salient characteristics such as age and socio-economic status. The intent is to reduce the extent to which differences in outcomes between the two groups reflect differences between them other than whether or not they received the intervention.

  4. A randomized trial is one in which individuals from the same population pool, such as all eligible appli- cants for a targeted intervention program, are random- ly assigned to the intervention or nonintervention group. This is considered to be the “gold standard” for any intervention evaluation, since it minimizes the possibility that differences in outcome will reflect something other than the effect of the intervention.

  5. This design compares two groups of children whose application was accepted by the program by using the age cut-off for enrolment eligibility to define the groups, one that had recently completed prekinder- garten and the other a group of children about to begin the program at the same time. The concept may be easier to understand by taking the extreme case, that of two children who differ only in that one was born the day before the date of cut-off and the other born the day after. When both are about to turn age five, the slightly younger child will be about to enter prekindergarten while the other child will be about to enter kindergarten. If the two children’s achievement scores are assessed on the same day, their closeness in age will mean that their maturation levels are virtually the same with the result that any differences in achievement scores reflect the effect of the prekinder- garten program. In practice, the approach is applied to a wider age range around the cut-off.

  6. Four studies measured school-readiness with the Early Development Instrument (EDI) which was developed by a consortium of Canadian researchers and early childhood experts in the late 1990s and has subsequently been val- idated (Janus and Offord 2007). The fifth study used the Peabody Picture Vocabulary Test — Revised as its meas- ure (Dunn and Dunn 1981).

  7. The EDI measures five components of school-readiness: physical health and well-being; social knowledge and competence; emotional maturity and ability to self-regu- late; language and cognitive development; and general knowledge and communication skills.

  8. ECEC quality was measured by trained observers while the child was in his or her ECE program using the Early Childhood Environment Rating Scale — Revised (Harms, Clifford and Cryer 1998) and the Caregiver Interaction Scale (Arnett 1989). A further discussion of what is meant by “quality” is provided later in this section.

  9. Prince Edward Island is an exception. In that province, practitioners with a two-year college ECEC credential are permitted to work in either child care or kindergarten and some child care centres deliver the kindergarten cur- riculum to age-eligible children as part of their child care program (Friendly et al. in press).

  10. The expenditure on ECEC calculated for each of the 14 countries compared was derived from information provid- ed by each country. Given the strong provincial/territorial role in child care in Canada, and the role of municipalities in Ontario, the information provided to the OECD presum- ably included funds from all three levels of government. 

Abbott-Shim, Martha, Robert Lambert, and Frances McCarty. 2003. “A Comparison of School Readiness Outcomes for Children Randomly Assigned to a Head Start Program and the Program’s Wait List.” Journal of Education for Students Placed at Risk 8 (2): 191-214.

ABC Head Start. 2006a. Annual Report, 2005-2006. Edmonton: ABC Head Start. Accessed February 6, 2007. www.abcheadstart.org

_____. 2006b. News n’ Views, Fall 2006. Edmonton: ABC Head Start. Accessed February 6, 2007. www.abcheadstart.org

Ackerman, Debra J., and W. Steven Barnett. 2006. Increasing the Effectiveness of Preschool Programs. Preschool Policy Brief 11. New Brunswick, NJ: National Institute for Early Education Research. Accessed March 21, 2007. http://www. nieer.org/resources/research/IncreasingEffectiveness.pdf

Administration for Children and Families. 2006. Preliminary Findings from the Early Head Start Prekindergarten Followup. Washington, DC: US Department of Health and Human Services. Accessed April 2, 2007. http://www.acf.hhs.gov/programs/opre/ehs/ehs_resrch/ index.html

Anderson, Laurie M., Carolynne Shinn, Mindy T. Fullilove, Susan C. Scrimshaw, Jonathan E. Fielding, Jacques Normand, Vilma G. Carand-Kulis, and the Task Force of Community Preventive Services. 2003. “The Effectiveness of Early Childhood Development Programs: A Systematic Review.” American Journal of Preventive Medicine 24:32-46.

Andersson, Bengt-Erik. 1992. “Effects of Day Care on Cognitive and Socioemotional Development of Thirteen-Year-Old Swedish Children.” Child Development 63:20-36.

Aos, Steve, Roxanne Lieb, Jim Mayfield, Marna Miller, and Annie Pennucci. 2004. Benefits and Costs of Prevention and Early Intervention Programs for Youth. Olympia, WA: Washington State Institute for Public Policy. Accessed February 19, 2007. http://www.wsipp. wa.gov/rptfiles/04-07-3901a.pdf

Arnett, Jeffrey. 1989. “Caregivers in Day Care Centers: Does Training Matter?” Journal of Applied Developmental Psychology 10:541-52.

Baker, Amy J.L., Chaya Piotrkowski, and Jeanne Brooks- Gunn. 1999. “The Home Instruction Program for Preschool Youngsters (HIPPY).” Home Visiting: Recent Program Evaluation 9 (1):116-33.

Baker, Michael, Jonathan Gruber, and Kevin Milligan. 2006. “What Can We Learn from Quebec’s Universal Child Care Program?” C.D. Howe Institute E-Brief, February. Accessed November 10, 2006. www.cdhowe.org/pdf/ ebrief_25_english.pdf

Barnett, W. Steven, 2002. Early Childhood Education. Greenwich, CT: Information Age Publishing.

Barnett, W. Steven, Kirsty Brown, and Rima Shore. 2004. “The Universal vs. Targeted Debate: Should the United States Have Preschool for All?” NIEER Preschool Matters 6. New Brunswick, NJ: National Institute for Early Education Research. Accessed February 17, 2007. www.nieer.org

Barnett, W. Steven, Jason T. Hustedt, Laura E. Hawkinson, and Kenneth B. Robin. 2006. The State of Preschool 2006. New Brunswick, NJ: National Institute for Early Education Research.

Barnett, W. Steven, Cynthia Lamy, and Kwanghee Jung. 2005. The Effects of State Pre-kindergarten Programs on Young Children’s School Readiness in Five States. New Brunswick, NJ: National Institute for Early Education Research. Accessed March 19, 2007. http://nieer.org/ resources/research/multistate/fullreport.pdf

Beach, Jane, Jane Bertrand, Barry Forer, Donna Michal, and Jocelyne Tougas. 2004. Working for Change: Canada’s Child Care Workforce. Ottawa: Child Care Human Resources Sector Council.

Beaudoin, André, and Daniel Turcotte. 2002. “Community Action for Children at Risk in Quebec.” In National and Regional Evaluations of the Community Action Plan for Children (CAPC): 1995-2000. Ottawa: Public Health Agency of Canada.

Belfield, Clive R. 2005. An Economic Analysis of Pre-K in Louisiana. Washington, DC: Pre-K Now. Accessed February 17, 2007. http://www.preknow.org/ documents/LAEconAnalysisReport_June2005.pdf

_____. 2006a. Does It Pay to Invest in Preschool for All? Analyzing Return-on-Investment in Three States. New Brunswick, NJ: National Institute for Early Education Research. Accessed November 29, 2007. http://nieer.org/resources/research/doesitpay.pdf

_____. 2006b. An Economic Analysis of Pre-K in Arkansas. Washington, DC: Pre-K Now. Accessed February 17, 2007. http://www.preknow.org/documents/ AREconAnalysisReport_Nov2006.pdf

Berlinski, Samuel, Sebastian Galiani, and Paul Gertler. 2006. The Effect of Pre-primary Education on Primary School Performance. London, UK: Institute for Fiscal Studies, Department of Economics, University College London. Accessed February 29, 2007. http://www.ifs.org.uk/wps/wp0604.pdf

Boyle, Michael H., and J. Douglas Willms. 2002. “Impact Evaluation of a National Community-Based Program for At-Risk Children in Canada.” Canadian Public Policy 28 (3): 461-81.

Bradley Robert, and Barbara Gilkey. 2002. “The Impact of the Home Instructional Program for Preschool Youngsters (HIPPY) on School Performance in 3rd and 6th Grades.” Early Education and Development 13 (3):302-11.

Brigance, Albert H. 1998. Preschool Screen, Kindergarten Screen and First Grade Screen. North Billerica, MA: Curriculum Associates.

Broberg, Anders G., Holger Wessels, Michael E. Lamb, and C. Phillip Hwang. 1997. “Effects of Day Care on the Development of Cognitive Abilities in 8-Year-Olds: A Longitudinal Study.” Developmental Psychology 33 (1): 62-9.

Brooks-Gunn, Jeanne, Greg J. Duncan, and Peter Britto. 1999. “Are Socio-Economic Gradients for Children Similar to Those for Adults? Achievement and Health of Children in the United States.” In Developmental Health and the Wealth of Nations, edited by Dan Keating and Clyde Hertzman. New York: Guildford Press.

Brownell, Marni, Noralou Roos, Randy Fransoo et al. 2006. “Is the Glass Half Empty? A Population-Based Perspective on Socioeconomic Status and Educational Outcomes.” IRPP Choices 12 (5).

Burchinal, Margaret, Sharon Landesman Ramey, Mary K. Reid, and James Jaccard. 1995. “Early Child Care Experiences and Their Association with Family and Child Characteristics during Middle Childhood.” Early Childhood Research Quarterly 10:33-61.

Bushnik, Tracey. 2006. Child Care in Canada. Ottawa: Statistics Canada, cat. no. 89-599-MIE2006003. Accessed March 5, 2007. http://www.statcan.ca/bsolc/ english/bsolc?catno=89-599-M2006003

Cameron, Claire. 2004. Building an Integrated Workforce for a Long-Term Vision of Universal Education and Care. London, UK: Day Care Trust. Accessed May 12, 2007. http://www.daycaretrust.org.uk/mod/ fileman/files/Building_an_Integrated_Workforce_ Final_28Nov04.pdf

Campaign 2000. 2006. Oh Canada! Too Many Children in Poverty for Too Long: 2006 Report Card on Child and Family Poverty. Toronto: Campaign 2000.

Canadian Council on Learning. 2006. Report on Learning in Canada, 2006, Canadian Post-Secondary Education: A Positive Record — An Uncertain Future. Toronto: Canadian Council on Learning. Accessed February 15, 2007. www.ccl-cca.ca

_____. 2007. State of Learning in Canada: No Time for Complacency. Toronto: Canadian Council on Learning. Accessed February 8, 2007. www.ccl-cca.ca

Canadian Council on Social Development. 2007. Urban Poverty Project 2007: Poverty Data Tables. Ottawa: Author. Accessed September 25, 2007. http://www.ccsd.ca/pubs/2007/upp/poverty_data_table s/provinces/canada.pdf

Canadian Intergovernmental Conference Secretariat. 2000. “First Ministers’ Meeting Communiqué on Early Childhood Development.” Press release, September 11. Ottawa: Canadian Intergovernmental Conference Secretariat.

Caughy, Margaret O’Brien, Janet A. DiPietro, and Donna M. Strobino. 1994. “Day-Care Participation as a Protective Factor in the Cognitive Development of Low-Income Children. Child Development 65:457-71.

Centraide of Greater Montreal. 2007. The 1,2,3 Go! Initiative. Montreal: Centraide. Accessed August 28, 2007. http://www.centraide-mtl.org/centraide/ static/where/go.shtml

Clements, Melissa A., Arthur J. Reynolds, and Edmond Hickey. 2004. “Site-Level Predictors of Children’s School and Social Competence in the Chicago Child-Parent Centers.” Early Childhood Research Quarterly 19:273-96.

Cleveland, Gordon, and Michael Krashinsky. 1998. The Benefits and Costs of Good Child Care: The Economic Rationale for Public Investment in Young Children. Toronto: Childcare Resource and Research Unit, University of Toronto.

Committee for Economic Development (CED). 2006. The Economic Promise of Investing in High-Quality Preschool: Using Early Education to Improve Economic Growth and the Fiscal Sustainability of States and the Nation. Washington, DC: CED. Accessed April 24, 2007. www.ced.org

Currie, Janet. 2005. Economic Impact of Head Start. Montreal: Centre of Excellence for Early Childhood Development. Accessed February 18, 2007. www.excellence-earlychildhood.ca

Currie, Janet, and Duncan Thomas. 1995. Does Head Start Make a Difference? The American Economic Review 85 (4): 341-64.

_____. 2000. School Quality and Longer-Term Effects of Head Start. Journal of Human Resources. 35 (4): 755-74. Dodge, David. 2003. “Human Capital, Early Childhood

Development and Economic Growth: An Economist’s Perspective.” Address to the Annual Sparrow Lake Alliance Meeting, May.

Doherty, Gillian. 1991. Addressing the Issue of Lack of School-Readiness in Preschoolers. Ottawa: Prosperity Steering Group, Government of Canada.

_____. 2001. Targeting Early Childhood Care and Education: Myths and Realities. Toronto: Childcare Resource and Research Unit, University of Toronto.

_____. 2007. Conception to Age Six: The Foundation Of School-Readiness. Toronto: Learning Partnership. Accessed March 10, 2007. http://www.thelearning partnership.ca/policy_research/Early_Years_research- paper-jan07.pdf

Doherty, Gillian. Barry Forer, Donna S. Lero, Hillel Goelman, and Annette LaGrange. 2006. “Predictors of Quality in Family Child Care.” Early Childhood Research Quarterly 21: 296-312.

Doherty, Gillian, Martha Friendly, and Jane Beach. 2003. OECD Thematic Review of Early Childhood Education and Care: Canadian Background Report. Ottawa: Her Majesty the Queen in Right of Canada.

Drouin, Carl, Nathalie Bigras, Claire Fournier, Hélène Desrosiers, and Stéphane Bernard. 2004. Grandir en qualité 2003: EnqueÌ‚te québecoise sur la qualité des serv- ices de garde educates. Quebec City: Institut de la statis- tique du Québec.

Dunn, Lloyd L., and Leota M. Dunn. 1981. Peabody Vocabulary Test — Revised. Circle Pines, MN: American Guidance Service.

Farran, Dale, C. 2000. “Another Decade of Intervention for Children Who Are Low Income or Disabled: What Do We Know Now?” In Handbook of Early Childhood Intervention, 2nd ed., edited by Jack P. Shonkoff and J. Meisels. New York: Cambridge University Press.

Federal-Provincial-Territorial Council of Ministers on Social Policy Renewal. 1999. A National Children’s Agenda: Developing a Shared Vision. Ottawa: Minister of Public Works and Government Services.

Fortin, Pierre. 2006. “Back to Basics.” Policy Options 27 (4). Montreal: IRPP.

Friendly, Martha, Jane Beach, Carolyn Ferns, and Michelle Turiano. In press. Early Childhood Education and Care in Canada 2006. Toronto: Childcare Resource and Research Unit.

Garces, Eliana, Duncan Thomas, and Janet Currie. 2002. “Longer-Term Effects of Head Start.” American Economic Review 92 (4): 999-1012.

Goelman, Hillel, Barry Forer, Paul Kershaw, Gillian Doherty, Donna S. Lero, and Annette LaGrange. 2006. “Towards a Predictive Model of Quality in Canadian Child Care Centers.” Early Childhood Research Quarterly 21:280-95.

Gomby, Deana. S. 2005. Home Visitation in 2005: Outcomes for Children and Parents. Washington, DC: Committee for Economic Development Invest in Kids Working Group. Accessed March 7, 2007. http://www.ced.org/docs/ report/report_ivk_gomby_2005.pdf

Goodson, Barbara D., Jean I. Layzer, Robert G. St. Pierre, Lawrence S. Bernstein, and Michael Lopez. 2000. “Effectiveness of a Comprehensive, Five-Year Family Support Program for Low-Income Children and Their Families: Findings from the Comprehensive Child Development Program.” Early Childhood Research Quarterly 15 (1): 5-36.

Gormley, William, T. 2006. Early Childhood Care and Education: Lessons and Puzzles. Georgetown, DC: Public Policy Institute, Georgetown University. Accessed February 19, 2007. http://gppi.georgetown.edu/pdfs/ Gormley%20Reading.pdf?CFID=302393&CFTO- KEN=26459712

Gormley, William T., Ted Gayer, Deborah Phillips, and Brittany Dawson. 2005. “The Effects of Universal Pre-K on Cognitive Development.” Developmental Psychology 41 (6):872-84. 

Government of Canada. 2004. Speech from the Throne to Open the First Session of the Thirty-Eighth Parliament of Canada (October 2004). Ottawa: Government of Canada

_____. 2005a. Federal Budget 2005. Ottawa: Department of Finance, Government of Canada.

_____. 2005b. Early Childhood Development and Expenditure; Early Learning and Child Care Activities and Expenditure, 2003-2004. Ottawa: Her Majesty the Queen in Right of Canada.

_____. 2006. Federal Budget 2006. Ottawa: Department of Finance, Government of Canada.

_____. 2007a. Federal Budget 2007. Ottawa: Department of Finance, Government of Canada.

_____. 2007b. Early Childhood Development and Expenditure; Early Learning and Child Care Activities and Expenditure, 2004-2005 and 2005-2006. Ottawa: Her Majesty the Queen in Right of Canada. Accessed August 30, 2007. http://www.socialunion.ca/ ecdelcc_ae/2007/en/a_e_report.pdf

Government of Manitoba. 2006. Annual Report of the Healthy Child Manitoba Office. Winnipeg: Healthy Child Manitoba. Accessed March 5, 2007. http://www.gov.mb.ca/healthychild/about/annual_ report_2005_06.pdf

Government of Quebec. 2007. Circumstances in Which the Standard $7 a Day Parent Contribution for Regulated Child Care May Be Reduced for Children Age 0 to 5. Accessed September 18, 2007. http://www.mfa.gouv.qc.ca/ thematiques/famille/services-garde/contribution- reduite/0-a-4-ans.asp

Government of Saskatchewan. 2005. KidsFirst Strategy: Annual Report, 2004-2005. Regina: Saskatchewan Learning. Accessed February 12, 2007. http://www.sasked.gov.sk.ca/branches/elcc/pdfs/ ECD_KidsFirst_AR04_05.pdf

_____. 2006. 2006-2007 Saskatchewan Provincial Budget, Performance Plan, KidsFirst Strategy. Regina: Minister of Learning, Minister of Healthy Living Services. Accessed February 26, 2007. www.sasked.gov.sk.ca

Harms, Thelma, Richard M. Clifford, and Debby Cryer. 1998. Early Childhood Environment Rating Scale, Revised Edition. New York: Teachers College Press.

Heckman, James. 2000. “Policies to Foster Human Capital.” Research in Economics 54 (1): 3-56.

Helburn, Suzanne. 1995. Cost, Quality and Child Outcomes in Child Care Centers. Denver, CO: Economics Department, University of Colorado at Denver.

Home Instruction for Parents of Preschool Youngsters (HIPPY) Canada. 2006. Focus Forward: Advancing 5 Key Initiatives, 2005-06 Annual Report. Vancouver: HIPPY Canada.

_____. Undated. How HIPPY Works. Vancouver: HIPPY Canada. Accessed February 2, 2007. http://www.hippycanada.com/howhippyworks.php

Howes, Carollee. 1990. “Can the Age of Entry into Child Care and the Quality of Child Care Predict Adjustment in Kindergarten?” Developmental Psychology 26 (2): 292-303.

Human Resources Development Canada. 2003. Supporting Canada’s Children and Families. Federal-Provincial-Territorial Ministers Responsible for Social Services Meeting press release, March 13. Ottawa: Human Resources Development Canada.

Hustedt, Jason, T., and W. Steven Barnett. 2005. Head Start Policy. Montreal: Centre of Excellence for Early Childhood Development. Accessed April 4, 2007. www.excellence-earlychildhood.ca

Hustedt, Jason T., W. Steven Barnett, Kwanghee Jung, and Jessica Thomas. 2007. The Effects of the Arkansas Better Chance Program on Young Children’s School Readiness. New Brunswick, NJ: National Institute for Early Education Research (NIEER). Accessed March 19, 2007. http://nieer.org/docs/?DocID=158

Institut de la statistique. 2006. “Population par groupe d’aÌ‚ge, Canada et régions, 2006.” Gouvernement du Québec. Accessed November 27, 2008. http://www.stat.gouv.qc.ca/donstat/societe/ demographie/struc_poplt/104.htm

Janus, Magdalena, and Dan Offord. 2007. “Development and Psychometric Properties of the Early Development Instrument (EDI): A Measure of Children’s School Readiness.” Canadian Journal of Behavioral Science 39 (1): 1-22.

Japel, Christa, Richard Tremblay, and Sylvie CoÌ‚té. 2005. “Quality Counts! Assessing the Quality of Day Care Services Based on the Quebec Longitudinal Study of Child Development”. IRPP Choices 11 (5). Accessed March 12, 2007. http://www.irpp.org/indexe.htm

Karoly, Lynn A., and James H. Bigelow. 2005. The Economics of Investing in Universal Preschool Education in California. Santa Monica, CA: RAND Corporation.

Karoly, Lynn A., M. Rebecca Kilburn, and Jill S. Cannon. 2005. Early Childhood Intervention: Proven Results, Future Promise. Santa Monica, CA: RAND Corporation. Accessed March 7, 2007. http://www.rand.org/pubs/ monographs/2005/RAND_MG341.pdf

Keating, Dan, and Clyde Hertzman. 1999. Developmental Health and the Wealth of Nations: Social, Biological and Educational Dynamics. New York: Guilford.

Kershaw, Paul, Lori Irwin, Kate Trafford, and Clyde Hertzman. 2005. The British Columbia Atlas of Child Development. Vancouver: Human Early Learning Partnership, University of British Columbia.

Kohen, Dafna, Clyde Hertzman, and J. Douglas Willms. 2002. “The Importance of Quality Child Care.” In Vulnerable Children, edited by J. Douglas Willms. Edmonton: University of Alberta Press.

Krentz, Caroline, Twyla Mensch, and Beth Warkentin. 2004a. Summary and Recommendations of the Third Year of a 6-Year Longitudinal Study Examining the Effectiveness of the Pre-kindergarten Program in the Regina Public School Division No. 4. Regina: Saskatchewan Instructional Development and Research Unit, University of Regina.

_____. 2004b. Summary and Recommendations of the Fourth Year of a 6-Year Longitudinal Study Examining the Effectiveness of the Pre-kindergarten Program in the Regina Public School Division No. 4. Regina: Saskatchewan Instructional Development and Research

Unit, University of Regina.
_____. 2005. Summary and Recommendations of the Fifth

Year of a 6-Year Longitudinal Study Examining the Effectiveness of the Pre-kindergarten Program in the Regina Public School Division No. 4. Regina: Saskatchewan Instructional Development and Research Unit, University of Regina.

_____. 2006. Summary and Recommendations of the Sixth Year of a 6-Year Longitudinal Study Examining the Effectiveness of the Pre-kindergarten Program in the Regina Public School Division No. 4. Regina: Saskatchewan Instructional Development and Research Unit, University of Regina.

Lamy, Cynthia, W. Steven Barnett, and Kwanghee Jung. 2005a. The Effects of the Michigan School Readiness Program on Young Children’s Abilities at Kindergarten Entry. New Brunswick, NJ: National Institute for Early Education Research (NIEER). Accessed March 19, 2007. http://nieer.org/resources/research/multistate/mi.pdf

_____. 2005b. The Effects of New Jersey’s Abbott Preschool Program on Young Children’s School Readiness. New Brunswick, NJ: National Institute for Early Education Research (NIEER). Accessed March 19, 2007. http://nieer.org/resources/research/multistate/nj.pdf

_____. 2005c. The Effects of South Carolina’s Early Childhood Programs on Young Children’s School Readiness. New Brunswick, NJ: National Institute for Early Education Research (NIEER). Accessed March 19, 2007. http://nieer.org/resources/research/multistate/sc.pdf

_____. 2005d. The Effects of Oklahoma’s Early Childhood Program on Young Children’s School Readiness. New Brunswick, NJ. National Institute for Early Education Research (NIEER). Accessed March 19, 2007. http://nieer.org/resources/research/multistate/wv.pdf

_____. 2005e. The Effects of West Virginia’s Early Childhood Program on Young Children’s School Readiness. New Brunswick, NJ. National Institute for Early Education Research (NIEER). Accessed March 19, 2007. http://nieer.org/resources/research/ multistate/wv.pdf

Lee, Valerie, and Susanna Loeb. 1995. “Where Do Head Start Attendees End Up? One Reason Why Preschool Effects Fade Out.” Educational Evaluation and Policy Analysis 17 (1): 62-82.

Lefebvre, Pierre, and Phillip Merrigan. 2003. “Assessing Family Policy in Canada, A New Deal for Families and Children.” Montreal: IRPP Choices 9 (5). Accessed February 12, 2007. http://www.irpp.org/fasttrak/index.htm.

LeMare, Lucy, and Karyn Audet. Undated. The Vancouver HIPPY Project: Preliminary Evaluation Findings from a Multicultural Program. Vancouver: Faculty of Education, Simon Fraser University.

Loeb, Susanna, Bruce Fuller, Sharon Lynn Kagan, and Bidemi Carrol. 2004. “Child Care in Poor Communities: Early Learning Effects of Type, Quality and Stability.” Child Development 75 (1): 47-65.

Love, John M., Ellen Eliason Kisher, Christine Ross, Jill Constantine, Kimberly Boller, Rachel Chazan-Cohen,

Christy Brady-Smith et al. 2005. “Effectiveness of Early Head Start for 3-Year-Old Children and Their Parents: Lessons for Policy and Programs.” Developmental Psychology 41 (6): 885-901.

Magnuson, Katherine A., Christopher Ruhm, and Jane Waldfogel. 2007. “Does Prekindergarten Improve School Preparation and Performance?” Economics of Education Review 26:33-51.

Marshall, Katherine, and Vincent Ferrao. 2007. “Participation of Older Workers.” Perspectives on Labour and Income 19 (3): 5-11. Accessed October 19, 2007. http://www.stat- can.ca/english/freepub/75-001-XIE/2007108/ articles/10303-en.pdf

McCain, Margaret Norrie, and J. Fraser Mustard. 1999. Early Years Study, Final Report. Toronto: Publications Ontario.

McCain, Margaret Norrie, J. Fraser Mustard, and Stuart Shanker. 2007. The Early Years Study 2: Putting Science into Action. Toronto: Council for Early Child Development. Accessed September 3, 2007. http://www.councilecd.ca.

Merrell, Kenneth. 2002. School Social Behavior Scale. 2nd ed. Eugene, OR: Assessment Intervention Resources. Accessed April 17, 2007. www.assessment-intervention.com

National Institute for Child Health and Human Development [NICHD] Early Child Care Research Network. 2002. “Early Child Care and Children’s Development Prior to School Entry: Results from the NICHD Study of Early Child Care.” American Educational Research Journal 39 (1): 133-64.

_____. 2005. “Early Child Care and Children’s Development in the Primary Grades: Follow-up Results from the NICHD Study of Early Child Care.” American Educational Research Journal 42 (3): 537-70.

_____. 2007. “Are There Long-Term Effects of Early Child Care?” Child Development 78 (2): 537-70.

NICHD Early Child Care Research Network and Greg J. Duncan. 2003. “Modeling the Impacts of Child Care Quality on Children’s Preschool Cognitive Development.” Child Development 74 (5): 1454-75.

National Institute for Early Education Research (NIEER). 2003. “Investing in Head Start Teachers.” Preschool Policy Matters 4. New Brunswick, NJ: NIEER. Accessed March 21, 2007. http://nieer.org/resources/policybriefs/4.pdf

Office of Planning, Research and Evaluation. 2007. Head Start Impact Study. Washington, DC: US Department of Health and Human Services. Accessed February 12, 2007. www.acf.hhs.gov

Olds, David L., Charles R. Henderson Jr., Robert E. Cole, John J. Eckenrode, Harriet J. Kitzman, Dennis Luckey, Lisa Pettitt, Kimberley Sidora, Pamela Morris, and Jane Powers. 1998. “Long-Term Effects of Nurse Home Visitation on Children’s Criminal and Antisocial Behavior.” Journal of the American Medical Association 280 (14): 1238-44.

Olds, David L., Charles R. Henderson Jr., Harriet J. Kitzman, John J. Eckenrode, Robert E. Cole, and Robert C. Tatelbaum. 1999. “Prenatal and Infancy Visitation by Nurses: Recent Findings.” The Future of Children 9 (1): 44-65.

Olds, David L., Harriet J. Kitzman, Robert E. Cole, JoAnne Robinson, Kimberly Sidora, Dennis W. Luckey, Charles R. 

Henderson Jr., Carole Hanks, Jessica Bondy, and John Holmberg. 2004a. “Effects of Nurse Home-Visiting on Maternal Life Course and Child Development: Age 6 Follow-up Results of a Randomized Trial.” Pediatrics 114 (6): 1550-9.

Olds, David L., JoAnn Robinson, Lisa Pettitt, Denis W. Luckey, John Holmberg, Rosanna K. Ng, Kathy Isacks, Karen Sheff, and Charles R. Henderson Jr. 2004b. “Effects of Home Visits by Paraprofessionals and by Nurses: Age 4 Follow-up Results of a Randomized Trial.” Pediatrics 114 (6): 1560-8.

Ontario. Ministry of Children and Youth Services. 2007a. “Ontario’s Best Start Plan.” Press release, July 5. Toronto. Ministry of Children and Youth Services.

_____. 2007b. Investing in Quality: Policies, Practitioners, Programs and Parents. Toronto: Ministry of Children and Youth Services. Accessed September 25, 2007. http://www.children.gov.on.ca/NR/CS/Publications/ QHRReport_en.pdf

Organisation for Economic Co-operation and Development (OECD). 2004. Early Childhood Education and Care Policy: Canada, Country Note. Paris: OECD Directorate for Education.

_____. 2006. Starting Strong II: Early Childhood Education and Care. Paris: OECD Directorate for Education. Osborn, A.F., and J.E. Milbank. 1987. The Effects of Early

Education: A Report from the Child Health and

Education Study. Oxford: Clarendon Press. Palacio-Quintin, Ercila. 2002. “National Impact

Evaluation Analysis.” In National and Regional Evaluations of the Community Action Plan for Children (CAPC): 1995-2000. Ottawa: Public Health Agency of Canada. 70-83.

Parent-Child Mother Goose Program. 2007. “About Us.” Toronto: Parent-Mother Goose Program. Accessed August 28 2007. http://www.nald.ca/ mothergooseprogram/About.htm

Peisner-Feinberg, Ellen S., and Margaret R. Burchinal. 1997. “Relations between Preschool Children’s Child-Care Experiences and Concurrent Development: The Cost, Quality, and Outcomes Study.” Merrill-Palmer Quarterly 43 (3): 451-77.

Peisner-Feinberg, Ellen S., Margaret R. Burchinal, Richard, M. Clifford, Mary L. Culkin, Carollee Howes, Sharon Lynn Kagan, and Noreen Yzzejian. 2001. “The Relation of Preschool Child-Care Quality to Children’s Cognitive and Social Developmental Trajectories through Second Grade.” Child Development 72 (5): 1534-53.

Peters, Ray V., Robert Arnold, Kelly Petrunka, Douglas Angus, Kathleen Brophy, Sharon Burke, Gary Cameron, Susan Evers, Yves Herry, Denis Levesque, Mark Pancer, Gloria Roberts-Fiati, Shelagh Towson, and Wendy Warren. 2000. Developing Capacity and Competence in the Better Beginnings, Better Futures Communities: Short-Term Findings Report. Kingston, ON: Better Futures Research Coordination Unit, Queen’s University.

Public Health Agency of Canada. 2004. Head Start Program Overview. Ottawa: Public Health Agency of Canada. Accessed May 13, 2007. www.phas-aspc.gc.ca.

_____. 2006. Community Action Plan for Children (CAPC), National Program Profile (NPP), Reporting Period: 2004-2005. Ottawa: Public Health Agency of Canada.

Puma, Michael, Stephen Bell, Ronna Cook, Camilla Heid, and Michael Lopes. 2005. Head Start Impact Study: First Year Findings, Executive Summary. Washington DC: Administration for Children and Families, US Department of Health and Human Services. Accessed February 12, 2007. http://www.acf.hhs.gov./ programs/opre/hs/impact_study/reports/first_yr_finds/ firstyr_finds_exec_summ.html

Quebec. Ministère de l’Éducation. 1996. A New Direction for Success: Ministerial Plan of Action for the Reform of the Educational System. Quebec City: Ministère de l’Éducation. Accessed February 12, 2007. http://www.mels.gouv.qc.ca/reforme/reform.htm

Québec. Ministère de l’Éducation, du Loisir et du Sport. 2007. Règles budgétaires pour l’année scolaire 2006- 2007: Commissions scolaires (Version amendée). Direction générale du financement et de l’équipement.

Quebec. Ministère de la Famille, des AiÌ‚nés et de la Condition feminine. 2007. Places en services de garde. Quebec: Ministère de la Famille, des AiÌ‚nés et de la Condition feminine. Accessed October 10, 2007. http://www.mfa. gouv.qc.ca/ statistiques/services-de-garde/index.asp

Quebec. Ministère de la Santé et des Services sociaux. 2004. Les services intégrés en périnatalité et pour la petite enfance à l’intention des familles vivant en contexte de vulnérabilité: Cadre de référence. Quebec: Ministère de la Santé et des Services sociaux. Accessed September 3, 2007. http://publications.msss.gouv. qc.caa/acrobat/f/documentation/2004/04-836-02W.pdf

Ramey, Craig T., Donna M. Bryant, Joseph J. Sparling, and Barbara H. Wasik. 1985. “Project CARE: A Comparison of Two Early Intervention Strategies to Prevent Retarded Development.” Topics in Early Childhood Special Education 5 (2): 12-25.

Ramey, Craig T., Sharon L. Ramey, K. Robin Gaines, and Clancy Blair. 1995. “Two-Generation Early Intervention Programs: A Child Development Perspective.” In Two Generation Programs for Families in Poverty: A New Intervention Strategy, edited by Sheila Smith. Norwood, NJ: Ablex Publishing.

Reynolds, Arthur J. 1994. “Effects of a Preschool Plus Follow-on Intervention for Children at Risk.” Developmental Psychology 30 (6): 787-804.

Reynolds, Arthur J., Wendy T. Miedel, and Emily A. Mann. 2000. “Innovations in Early Intervention for Children in Families with Low Incomes — Lessons from the Chicago Child-Parent Centers.” Young Children 55 (2): 84-8.

Reynolds, Arthur J., Judy A. Temple, Dylan L. Robertson, and Emily A. Mann. 2002. “Age 21 Cost-Benefit Analysis of the Title I Chicago Child-Parent Centers.” Educational Evaluation and Policy Analysis 24 (4): 267-303.

St. Pierre, Robert G., Jean I. Layzer, and Helen V. Barnes. 1998. “Regenerating Two-Generation Programs.” In Early Care and Education for Children in Poverty: Promises, Programs and Long-Term Results, edited by W. Steven Barnett and Sarane S. Boocock. Albany, NY: SUNY Press. 

Sammons, Pam, Kathy Sylva, Edward Melhuish, Iram Siraj- Blatchford, and Brenda Taggart. 2007. The Effective Preschool and Primary Education 3-11 Project (EPPE 3-11), Summary Report, Influences on Children’s Attainment and Progress in Key Stage 2: Cognitive Outcomes in Year 5. London, UK: Department of Education and Skills.

Sammons, Pam, Kathy Sylva, Edward Melhuish, Iram Siraj- Blatchford, Brenda Taggart, and Karen Elliot. 2002. The Effective Provision of Preschool Education Project: Technical Paper 8a — Measuring the Impact of Preschool on Children’s Cognitive Progress over the Preschool Period. London, UK: Department of Education and Skills.

Saskatchewan Learning. 2006. Early Learning and Child Care Prekindergarten Program. Accessed February 9, 2007. http://www.sasked.gov.sk.ca

Statistics Canada. 2007. Women in Canada: Work Chapter Updates 2006. Ottawa: Ministry of Industry. Accessed May 4, 2007. http://www.statcan.ca/english/ freepub/89F0133XIE/89F0133XIE2006000.pdf

Stipek, Deborah. 2001. “Pathways to Constructive Lives: The Importance of Early School Success.” In Constructive and Destructive Behavior: Implications for Family, School and Society, edited by Deborah Stipek and Arthur Bohart. Washington, DC: American Psychological Association.

Sylva, Kathy, Edward Melhuish, Pam Sammons, Iram Siraj- Blatchford, and Brenda Taggart. 2004. The Effective Provision of Preschool Education (EPPE) Project: Final Report. London, UK: Department of Education and Skills.

Tougas, Jocelyne. 2002. Reforming Québec’s Early Childhood Care and Education: The First Five Years. Toronto: Childcare Resource and Research Unit, University of Toronto.

Vandell, Deborah Lowe. 2004. “Early Child Care: The Known and the Unknown.” Merrill-Palmer Quarterly 50 (3): 387-414.

Vandell, Deborah Lowe, and Mary Anne Corasaniti. 1990. “Variations in Early Child Care: Do They Predict Subsequent Social, Emotional and Cognitive Differences?” Early Childhood Research Quarterly 5: 555-72.

Vandell, Deborah Lowe, and Barbara Wolfe. 2000. Child Care Quality: Does It Matter and Does It Need to Be Improved? Washington, DC: US Department of Health and Human Services.

Wagner, Mary M., and Serena L. Clayton. 1999. “The Parents as Teachers Program: Results from Two Demonstrations.” Home Visiting: Recent Program Evaluations 9 (1): 91-115.

Wechsler, David. 2001. Wechsler Individual Achievement Test II (Abbreviated). Toronto: Psychological Corporation-Harcourt Assessment.

Western Arctic Aboriginal Head Start Council (WAAHSC). 2007. Ten Years of Aboriginal Head Start in the NWT. Yellowknife, NWT: WAAHSC.

Whitebook, M. 2003. Bachelor’s Degrees Are Best: Higher Qualifications for Pre-kindergarten Teachers Lead to Better Learning Environments for Children. Washington, DC: Trust for Early Education.

Willms, Douglas J. 2002. “The Prevalence of Vulnerable Children.” In Vulnerable Children: Findings from Canada’s National Longitudinal Study of Children and Youth, edited by J. Douglas Willms. Edmonton: University of Alberta Press.

Winnipeg School Division. 2006. Nursery Program. Winnipeg: Winnipeg School Division. Accessed February 9, 2007. http://ww.wsd1.org/parents/documents/ NurseryProgram.pdf

Wylie, Cathy, Edith Hodgen, Hilary Ferral, and Jean Thompson. 2006. Contributions of Early Childhood Education to Age-14 Performance: Evidence from the Competent Children, Competent Learners Project. Wellington, NZ: Ministry of Education. Accessed February 7, 2007. www.nzcer.org.nz.

Yau, Maria Y.M. 2000. Do In-School Parenting and Family Literacy Centres Make a Difference? Findings from First Year Evaluation, 1999-2000. Toronto: Toronto District School Board.

_____. 2005. “Do Parenting and Family Literacy Centres Make a Difference?” Research Today 1 (1): 1-4. Toronto District School Board.

Zigler, Edward F., and Sally Styfco. 1996. “Head Start and Early Childhood Intervention: The Changing Course of Social Science and Social Policy.” In Children, Families and Government: Preparing for the 21st Century, edited by Edward F. Zigler and Sally Styfco. London, UK: Cambridge University Press.

Zill, Nicholas, Gary Resnick, Kwang Kim, Kevin O’Donnell, Alberto Sorongon, Ruth Hubel McKey, Shefali Pai- Samant, Cheryl Clark, and Mary Ann D’Elio. 2003. Head Start FACES 2000: A Whole-Child Perspective on Program Performance. Fourth Progress Report. Washington, DC: Administration for Children and Families, US Department of Health and Human Services. Accessed April 27, 2007. http://www.acf.hhs.gov/ programs/opre/hs/faces/reports/faces00_4thprogress/ faces00_4thprogress.pdf