Directorates
·	Director General's Office
·	Data Development
·	Strategic Evaluations
·	Partnership and Continuous Evaluations
Publications
·	Evaluation Reports
·	Evaluation Briefs
·	Labour Market Development Agreement Series
·	Lessons Learned Series
·	Evaluation Tool Kit Series
·	Evaluation Forum Newsletter
·	From Theory To Practice

Direct Job Creation Programs:
Evaluation Lessons

Arun S. Roy and
Ging Wong

Evaluation and Data Development
Strategic Evaluation and Monitoring
Strategic Policy
Human Resources Development Canada

December 1998

SP-AH082E-12-98

Parts of this study were presented at the HRDC National Policy Seminar on Job Creation, Toronto, May 1996. Authors are grateful to the participants and discussants for their encouragement, comments and suggestions.

Canada has experimented with a variety of direct job creation programs over the past twenty-five years. Although many of the programs introduced by the federal government have been evaluated from time to time, it is difficult to make comparisons across programs. For example, the available evaluation results are scattered over numerous reports. Also, the evaluation studies have not followed a consistent methodology, and the available program data do not always match across programs.

This evaluation study provides a review of the direct job creation programs introduced by the federal government. The review brings diverse evaluation findings into a common focus. It also draws evaluation lessons about what works and what does not work.

To help make program comparisons, the review develops a measure of the incremental cost of job creation for each program. The measure incorporates available information on the incremental employment impacts of each program and the average duration of the jobs.

The review finds that the cost effectiveness of direct job creation programs improved considerably over time as program design was continuously modified, ineffective elements were dropped, and better elements were added. Also, the lessons drawn from the available evaluations highlight important design issues and the potential contributions and limitations of these types of programs.

Like many other OECD countries, Canada has experimented with a variety of job creation programs over the last twenty-five years.These can be loosely classified as “direct” and ‘indirect” job creation measures, even though the distinction may often appear to be somewhat blurred.

• The main emphasis of direct job creation measures is to create jobs that are of a short-term nature to reduce the number of unemployed, reduce dependence on social programs, and to some extent stimulate the economy to create jobs on a longer term basis. Canada Works and the Local Initiatives Program would be typical examples of direct job creation programs in the Canadian context.

• Indirect job creation programs are aimed at improving the employability of job-seekers and increasing the level of long-term employment with greater emphasis on training and skill development. The Employer Wage Subsidy and Self-Employment Incentive\Assistance programs are examples of indirect job creation programs.

The purpose of this study is to provide a brief review of direct job creation programs and draw some lessons learned on the basis of evaluation studies that have been done from time to time. The evaluation studies have not followed consistent methodology and the findings are scattered in numerous reports. Therefore, this study attempts to bring the evaluation findings into a common focus with a view to drawing inferences as to what works and what does not work.

The direct job creation programs focused in this review generally correspond to the definition used by the OECD in its reports (OECD, 1980, 1997). The list of programs covered in this report does not, however, pretend to be exhaustive. Instead, the list was determined by two criteria. Some programs were selected for this review because they were large in terms of program expenditure. More often, however, programs were selected because they had a relatively long life and hence had been subjected to a formal program evaluation.

Unless otherwise stated, job creation in this report refers to federal direct job creation initiatives in the public sector and excludes direct job creation programs initiated by provinces.

It seems appropriate to point out that several difficulties are encountered in this kind of exercise.

• First, some programs were in operation only for a short period of time and were discontinued even before a formal evaluation was completed. In such cases, a complete evaluation report is unavailable and program data are rather sketchy.

• Second, when a new government came to power, quite often the main structure of an existing program was retained but the program title was changed giving the impression of a new job creation program.

• Third, the available data on the job duration of program participants and on program expenditures do not always match. For example, sometimes program participation data refers to weeks of employment during the calendar year, while program expenditures have been reported for the financial year. The lack of matching data makes it difficult to estimate the cost of job creation by program. For this reason, costs of job creation are presented in this report only for those programs where the calculation was possible and meaningful. In addition to information from past evaluation studies, program administration data and Departmental Annual Reports have been used to supplement statistical information.

Although details and emphasis tend to differ from program to program, there are some common features that run through the direct job creation programs reviewed in this report.

1.1  Objectives and Target Groups

The programs were aimed at one or more of the following:

1. reducing cyclical unemployment;
2. targeting certain disadvantaged labour force or demographic groups;
3. relieving certain structural labour market imbalances; and
4. providing services or producing goods of social value that would not otherwise be generated by the private sector or government agencies.

For instance, the Canada Works program targeted the generally unemployed for its participants. The New Employment Expansion and Development (NEED) program was focused on the long-term unemployed. Other programs targeted specific demographic groups such as youth (e.g. Young Canada Works), women, older workers, aboriginal people, etc.

1.2  Sponsors and Projects

Although these programs were initiated by a federal government agency, not all jobs were created in the public sector. Some programs, such as Canada Works and the Local Initiatives Program (LIP), gave priority to private non-profit organizations with the proviso: “projects should provide useful services or facilities to communities”. Due to this specific provision, a major part of the job creation was in highway construction, road repair, and environmental improvement. For example, two-thirds of the jobs created by Canada Works and LIP were in construction and environment-related projects. Projects were usually sponsored by municipal governments and community agencies, due to the additional restriction that these projects should not be profit-oriented but should promote public welfare.

1.3 Participants

Since the jobs were of a temporary nature, and were located in construction and related industries, the participants were generally low skilled with employment problems. However, depending upon the pool of unemployed available at the time, a program could attract participants with relatively high skills. For instance, about 25 percent of participants in Canada Works in 1977 did not have more than eight years of schooling but, at the other end of the educational spectrum, 26 percent had some post-secondary education or a college or university degree. In the case of a program, such as New Technology and Expansion Program (NTEP) which targeted new graduates in science and technology, the majority of program participants were university or college graduates.

Since many of the projects under Canada Works, LIP, and LEAP involved a fair amount of manual work, female participation levels were much lower than male participation levels. Many of these projects involved younger people more than prime-age workers. Also, the program participants were generally in low-wage jobs (at or marginally above minimum wage).

1.4  Organization

All federal job creation programs were designed and funded by a federal agency. In most cases, however, project selection and administration were left to local and municipal governments because local agencies had first hand knowledge of the unemployed and the projects that could be of value to the community. In the case of NEED, there was a tripartite partnership with contributions from private sector and the provinces.

1.5  Duration

The projects supported by the federal direct job creation programs were of a short-term nature. The time limit on participation in a program usually did not exceed 12 months.

In many cases, participants rarely stayed on a project for the maximum duration allowed by the program. For instance, Young Canada Works allowed a maximum of 14 weeks for a job but the average job duration was 10.5 weeks. Similarly, the maximum allowed job duration under the NEED program was specified at 52 weeks, but the average was 23 weeks.

Despite the intentions of program planners to the contrary, job creation programs tended to arouse the expectations of the community and employers. These expectations, in turn, tended to create a continuing dependence on publicly funded jobs or at least tended to generate pressures to extend the life of a program.

1.6  Program Logic Model

Figure 1 presents the Program Logic Model of direct job creation programs. It provides a description of how these programs generally work and how they impact on the economy.

The Program Logic Model identifies some of the major target groups: unemployed (on UI); social assistance recipients (SARs); employment-disadvantaged groups such as women, youth and aboriginals; and other groups such as older workers and new entrants into the work force.

The Program Logic Model also distinguishes between primary and secondary impacts (which in a broad sense can be used interchangeably with short-term and long-term impacts respectively). The Program Logic Model also indicates how a job-displacement effect, fiscal-substitution effect, and crowding-out effect could arise as a result of job creation initiatives.

View Figure 1 - Program Logic Model

When a job creation program is mounted, it could displace some workers that would have found jobs if the program intervention had not been made. This potential job-displacement effect is reflected in the concept and measurement of employment incrementality (see Section 2 – Evaluation Methodology and Data).

Some concerns are raised in program evaluation with regard to the fiscal-substitution effect of program expenditure. If a program is continued over a long period of time, it could be at the expense of some other potential program expenditure and hence there could be a fiscal-substitution effect. Some American evaluation studies have reported large fiscal-substitution effects of job creation programs (Fechter, p.17).

Sometimes concerns are also expressed about the crowding-out effect in the private sector. Since job creation programs imply boosting the demand for labour and creating employment, in theory these programs could drive up market wages and reduce employment for others. However, job creation programs usually target the unemployed; and therefore such crowding-out effects are unlikely to be of great concern. In addition, since direct job creation programs are usually directed to activities that are non-profit oriented and hence unlikely to be of interest to the private sector, such impacts would be minimal if they did occur. Several studies which have addressed this issue have reported nil or insignificant job-displacement effects in the private sector in other countries (OECD, p.31).

1.7  Short-term vs. Long-term Impacts

In the short-term, a direct job creation program takes individuals off welfare rolls and the UI system by providing new employment and, therefore, generates savings to the government. In the longer-run, the newly generated employment and income, with their secondary effects, are expected to increase government revenue. Also, after the participation has ended, these programs are expected to have improved the participants’ employability through work experience and training.

However, a direct job creation program may have several unintended impacts. It could qualify/re-qualify a UI claimant or increase the UI benefit period of the participants, thereby recycling the unemployed between a job creation program and UI. Also, new job creation initiatives could induce individuals in the non-labour force to enter the labour market and, therefore, add to the number of people reported as being unemployed.

This report focuses mainly on the cost of job creation as an evaluation measure. The cost per job is measured as the ratio of program expenditure over the number of jobs created. This provides only a crude measure of cost per job, however, because two other factors need to be considered and adjustments made. First, job duration varies by program; for example, some jobs may last for 10 weeks while others may last for 50 weeks or longer. Secondly, all jobs generated by a program may not be incremental. The employment incrementality (i.e. the number of jobs which would not have been created without the program intervention) also varies by program.

The standard and most acceptable methodology that is applied to estimate the incremental program impact consists of using a control or comparison group. To apply this kind of analysis, longitudinal data are needed for program participants as well as comparable non-participants¹. Since direct job creation programs are usually of short duration, however, precise measurement of program incrementality is not always possible.

A less rigorous method consists of self-assessed incrementality through sample surveys that ask program participants whether they would have found jobs if the program had not existed. Similarly, employers are asked whether they would have created the jobs without the program intervention. This measure of incrementality is necessarily subjective. However, the element of subjectivity can be reduced to some extent by designing the survey questionnaire from diverse angles² . The Canadian evaluation studies of direct job creation programs have relied on this approach.

The incremental cost of job creation (C) in this paper is to be interpreted as the incremental cost of job per work-year and is estimated as:

Where:

To help make comparisons over time, the estimates of the incremental cost per job in current dollars were converted into constant dollars. Data on incrementality ratios were obtained from the program evaluation studies. Also, data on program expenditure, number of jobs created, and the average duration of participation in a program were usually obtained from the evaluation studies. Where this information was not complete, program administration data and Departmental Annual Reports were used to supplement the data.

The following cautions should be noted in interpreting the cost of job creation estimates. First, the cost refers to the cost to the federal government only and does not include the expenses incurred by the employers or sponsors in creating and maintaining a job. Secondly, the cost refers to the program expenditure only and does not include the cost of program administration. Wherever the program data included program administration costs, such costs were taken out.

The measure of the incremental cost of job creation used in this study has several limitations that may be noted.

• The cost measure does not reflect the program externalities, which are often conceptualized in a comprehensive cost-benefit evaluation framework. Such program externalities would include the social opportunity cost of labour; secondary impacts in the form of income and employment multiplier effects at the community level; saving in UI and welfare payments for program participants; benefits to the government due to increased tax revenue resulting from incremental employment. Table 1 provides a schematic presentation of these elements. Needless to say, these elements of cost and benefit cannot be incorporated for job creation programs without many heroic assumptions³.

• The cost measure does not reflect the wage versus the non-wage components of a given program expenditure. Therefore, it tends to overstate the effectiveness of job creation programs that are oriented towards projects that are more labour intensive and less capital intensive. However, since direct job creation programs are usually labour intensive and wages constitute 90 to 95 percent of the program expenditure, this should not be a source of major concern.

• The present study takes the more modest approach of reporting on some rough measures of program incrementality and the incremental cost of job creation under different direct job creation programs with a view to drawing some tentative lessons for the future.

3.1  Expenditure on Job Creation Programs

Precise amounts spent on direct job creation programs are hard to come by and are difficult to construct. However, some recent OECD reports have filled this gap somewhat.

Recent information shows that public expenditure on labour market programs is low in Canada in comparison with other OECD countries (see Table 2). Canada spends only 19 currency units (dollars) out of every 1000 units (dollars) of GDP on labour market programs. By comparison, Sweden spends 43 currency units per 1000 currency units of GDP.

The amount spent in Canada on active measures, as a proportion of expenditure on all labour market programs, is closer to other OECD countries. For example, Canada spends 30 percent on active measures, in comparison with 34 percent and 33 percent respectively in Denmark and Belgium. However, spending on active measures in Sweden and the U.S. is 49 percent and 40 percent respectively.

Source: Compiled from OECD, Employment Outlook, July 1997, and July 1998, Tables K and J respectively.

¹ For example, $19 out of $1,000 of GDP is spent on labour market programs in Canada.

² Active labour market measures are defined to include public employment services and administration, labour market training, youth measures, subsidized employment, and measures for the disabled. By comparison, passive or inactive labour market measures consist mainly of unemployment insurance/compensation.

Table 3 presents public expenditure on direct job creation programs as a percent of public expenditure on all labour market programs. The amount spent on direct job creation programs is the lowest in Canada (only 1.6 percent) as compared with 2.3 percent in the U.S., 4.3 percent in Denmark, and 13.5 percent in Belgium. These results are also illustrated in Figure 2. A similar conclusion emerges from Table 4 and Figure 3, which present data on participation inflows in direct job creation as a ratio of the labour force.

Sources: Compiled from OECD, Employment Outlook, July 1997, and July 1998, Tables K and J respectively.

¹ See note under Table 2. Direct job creation measures refer to programs targeted at public and non-profit sector.

View Figure 2

View Figure 3

For Canada, the amounts spent on direct job creation programs were slightly higher in previous years: 2.3 percent and 1.8 percent in 1993-94 and 1994- 95 respectively, as compared with 1.6 percent in 1996-97. This downward trend in Canada probably reflects a shift in recent years from direct job creation programs towards developmental uses of UI as a means of job creation (see Sub-section 4.10 – Developmental Uses of UI). Canadian experience with direct job creation programs is compared with that of other countries at the end of this report (see Section 6 – Experience of Other Countries).

In Canada, direct job creation activity by the federal government passed through distinct phases. It is possible to divide the last twenty five-year period into three phases:

1. Pre-CJS (Canadian Jobs Strategy) Regime covering the period from 1971 to the first half of the eighties;

2. Canadian Jobs Strategy Regime extending roughly from 1985 to 1990; and

3. Labour Force Development Strategy/Employment Insurance Regime extending roughly from 1991 to the present time.

3.2 Pre-CJS (Canadian Jobs Strategy) Regime

During the 1970’s, thousands of young baby boomers were flooding the Canadian labour market to look for full-time jobs and particularly summer jobs. The youth unemployment rate was increasing (Figure 4). The government was anxious to avoid the social unrest caused by unemployed youth that was being experienced in other countries. Thus the early 1970’s witnessed a surge of new programs designed for students and youth.

View Figure 4

Some of the major job creation programs during this decade were: Opportunities For Youth (OFY), Local Initiatives Program (LIP), Local Employment Assistance Program (LEAP), Community Employment Strategy (CES), Canada Works (CW), Young Canada Works (YCW), Summer Job Corps (SJC), Co-Operative Education Program (COOP), and Employment Tax Credit Program (ETCP). In retrospect, the seventies can be described as the golden age of direct job creation. A brief description of the individual programs and their components is provided in Section 4.

During the eighties, although there were some important job creation initiatives, the most innovative ones were in the area of work sharing, which had the main objective of preserving existing jobs and assisting workers to adjust to major layoffs due to the recession of 1981-82.

During the first half of the eighties, in response to the recession, the Canada Works program was revived and experiments were conducted on the use of UI funds in support of job creation under Section 38 of the UI Act. In addition, several other new job creation programs were introduced: Canada Community Development Projects (CCDP), Canada Community Services Projects (CCSP), Local Economic Development Assistance (LEDA), Summer Youth Employment Program (SYEP), and Local Employment Assistance and Development (LEAD). Details on these programs are presented in Section 4.

3.3 Canadian Jobs Strategy (CJS) Regime

As the effects of the recession eased, Employment and Immigration Canada (EIC) began to implement the Dodge and Allmand reports and formulated a new philosophical framework for job creation and employment programming⁴. The new umbrella labour market policy, called Canadian Jobs Strategy (CJS) was announced in June, 1985. The basic principle was “the Canadian Jobs Strategy aims at ensuring that federal resources are used effectively to bring direct assistance to those most in need”⁵.

The notion of “most in need” recognized that there would never be enough resources to do everything worth doing. It implied that EIC would try to direct aid to workers with the most severe employment problems, areas with highest unemployment, and industrial sectors with the worst skill shortages. In calling for an effective use of federal resources, the principle also argued that the “the most in need” approach should be combined with “what works best” and with an increased emphasis on the market principle. The CJS regime witnessed a reduced emphasis on direct job creation programs and the introduction of indirect employment programs such as Job Entry, Job Development, Skill Investment, and Skill Shortages. The Community Futures program was also introduced. A major component was the Self-Employment Incentive (SEI), which was not aimed at direct job creation but at assisting workers to become self-employed.

3.4 Labour Force Development Strategy/Employment Insurance Regime

From about 1991, the traditional types of direct job creation programs (Canada Works, LIP, LEAP, CEP) were gradually phased out in favour of job creation under Developmental Uses of UI (see Section 4). The period can be described as the Labour Force Development Strategy/Employment Insurance Regime (LFDS/EI Regime).

Figure 5 illustrates how some of the major Canadian direct job creation programs evolved and were replaced by other programs over time. More details on these programs, their inter-relationships, and some evaluation results are presented in Subsections 4.1 to 4.11.

View Figure 5

4.1 Local Initiatives Program (1971-77)

The Local Initiative Program (LIP) was introduced in November of 1971. It had a two fold objective: (i) to create additional jobs during winter months from November to May of the following year; (ii) to invoke the participation and involvement of community groups and the unemployed to provide services that would benefit the whole community.

Projects were to provide at least 15 months of employment and a substantial part of it was to be additional employment over and above what would normally take place. Projects were to be of a non-profit nature. Project organizers were to use Canada Manpower Centres to hire unemployed workers.

Initially, $50 million was allocated for municipal projects and $50 million for privately sponsored projects. LIP operated for six winters, from 1971-77. In the program’s final winter, 1976-77, the original budget was increased to $190 million as unemployment rose sharply.

In 1974-75, a total of 4,150 projects were approved and the LIP contribution was $80 Million. These projects created a total of 31,160 jobs with an average duration of 20 weeks per job. The average gross cost per work-year was $6,675 (Manpower and Immigration, 1977). The incrementality ratio was 78 percent⁶ (i.e. 78 percent of employment created under the program would not have been created without the program).

LIP projects covered a broad range of activities from day-care centres to bridge building. Of those employed, about 25 percent were UI recipients and slightly more than another one-quarter were on social assistance. Women were under-represented in comparison with their share in the unemployed.

LIP was assessed to be an excellent program to create jobs quickly and at a low cost to the government. However, the Auditor General’s Report pointed out delays in data collection for evaluation and weaknesses in clarifying some program objectives in measurable terms. One example was “community betterment”, which was stated as an important goal of the program.

LIP was replaced in 1978 by a very similar program, called Canada Works.

4.2 Canada Works (1977-80)

Canada Works was introduced in 1977 as a successor to LIP. It was designed to create employment opportunities for unemployed Canadians by utilizing their skills for services to their communities. The objective was to get needed work done to the public benefit and by people whose energies and talents were temporarily surplus to private sector needs.

Funded at $253.6 million in 1977-78, Canada Works constituted the main element of that year’s employment strategy (EIC, 1980).

Wages for those who worked on the projects were generally about 10 percent above provincial minimum rates. For up to 52 weeks, projects could receive funding allocated on a constituency basis relative to the incidence of unemployment.

Ministerial Advisory Boards composed of community representatives were created in each constituency to review project proposals and make recommendations to the Minister who had final approval on projects.

The gross cost per job under Canada Works was estimated to be about $8,000 per work-year for 1978-79 in 1979 dollars and $10,000 for 1979- 80 in 1980 dollars (see EIC 1980, p.12). The average duration of participation was 25 weeks. The incrementality was estimated to be of the order of 80 to 90 percent (i.e. 80 to 90 percent of employment created under the program would not have been created without the program).

About two-thirds of all jobs created were in construction and environmental improvement projects. Priority was given to private non-profit organizations. Among participants, the mix of skills/educational background was rather heterogeneous. About one-quarter of the participants in 1977 did not have more than eight years of schooling; however, at the other end of the educational spectrum, 26 percent had some post-secondary education or a college or university degree.

An Economic Growth Component (ECG) feature was added to the Canada Works program in 1978-79 under which federal departments and agencies could submit job creation proposals. The eligibility criteria consisted of federal projects having the potential of contributing to economic growth and creating continuing employment in the private sector. The job creation had to be incremental and the project had to have a high probability of not otherwise occurring (EIC, April 1980). In addition, the project location was to be in constituencies with above-average unemployment (Cullen, June 1977, p.22). If a proposal was accepted, the funds required were transferred to the department or agency’s budget. In 1978-79, a total of $28.4 million was transferred and 3,163 work-years of employment were created (Hunter, undated).

The following are the findings and lessons that emerge from an evaluation of the EGC of the Canada Works program:

• The gross cost per work-year was $13,700 and was based on projects covering both years of ECG operation, that is 1978-79 and 1979-80 (EIC, April 1980). Thus, the cost of the ECG component was slightly higher than for the Canada Works in general. The estimated incrementality ratio for the ECG component was also somewhat lower at 78 percent compared to 80 to 90 percent for Canada Works in general.

• Based on an in depth analysis of some selected ECG projects, the lowest cost-benefit ratio was 0.63 and the highest was 5.0. The evaluation report noted that “the fundamental reason for such a wide variation in the results among projects relates to differences in the estimated probability of employment of project participants” (EIC, 1980).

• ECG experience suggests that the majority of federal departments do not consider job creation to fall within their mandate, and as such the response from them was rather poor.

• The project selection criteria and guidelines were vague and too much discretion was left with program officers. In addition, the accounting framework was weak and “there was no established authority to demand accounting for expenditures and adherence to program goals…”. Also, “Unemployment data did not play an obvious role in choosing projects…” (EIC, 1980, p.7-8). Further, inordinate delays in the project approval process was not conducive to program implementation.

• Adequate consideration was not given to data requirements and the data collection process for program evaluation.

Another special feature added in 1978-79, the Alternate Use of Canada Works Funds, allowed EIC to enter into four-year funding agreements with the provinces to create jobs in areas of provincial jurisdiction. In April 1979, responsibility for this feature was transferred to the Department of Regional Economic Expansion. Canada Works was replaced by CCDP (Canada Community Development Projects) in 1980.

4.3 Local Employment Assistance Program (1972-83)

Introduced in 1972, the Local Employment Assistance Program (LEAP) was designed to “create worthwhile employment for people who have been unemployed or receiving public assistance and who are not likely to become employed through normal labour market activity” (Manpower and Immigration, Annual Report, 1972-73, p.9).

The program was aimed at projects with the potential to become self-sufficient businesses. A large number of projects were targeted at native and disadvantaged groups including the physically and mentally handicapped, alcohol and drug abusers, and welfare recipients. In 1980-81, 48 percent of the LEAP participants were natives, 31 percent were women, 8 percent were handicapped, and 42 percent were youth including ex-offenders. By labour force status, 85 percent were unemployed; by income level, 64 percent had a low income level of less than $3,000; and regarding income source, over 30 percent were on welfare (EIC, 1982b).

Although initially started with a budget of less than $6 million in 1972-73, LEAP gradually expanded to $24 million by 1978-79. During 1980-81, LEAP created over 8,800 jobs, with an estimated gross cost of $10,600 per person year and an average employment duration of 33 weeks (EIC, 1982b). The incrementality ratio varied between 60 to 70 percent, based on a survey of project sponsors.

LEAP was seen as an effective means of helping some of the most disadvantaged groups of people in the labour market. Over 50 percent of LEAP participants found jobs within two months in the post-LEAP period.

LEAP was incorporated into Local Employment Assistance and Development (LEAD) in 1983.

4.4 Local Economic Development Assistance (1980-83)

Introduced in 1980, Local Economic Development Assistance (LEDA) was designed to play a direct and active role in stimulating private sector employment through local enterprise development within communities. LEDA funded the planning and operation of community-based corporations that would provide technical support and financial assistance to proposed or existing local businesses. The latter would then create continuing jobs in the community.

Jointly administered by EIC and the Department of Regional Economic Expansion, LEDA had a budget of only $3 million in 1980-81. Thus, the financial assistance component of the program was very limited. It created about 600 private sector jobs. The gross cost per job works out to $5,000 per job (not per work-year). The net cost could not be computed due data limitations.

The evaluation noted that LEDA corporations served a useful economic purpose in local communities and had been able to tap the expertise of local industrial/ economic development bodies. The report also noted that there was an over-representation of business people and an under-representation of participants from other groups.

In 1983, LEDA was incorporated into the LEAD (Local Employment Assistance and Development).

4.5 Canada Employment Program (1980-83)

Canada Employment Program (CEP) was introduced in 1980 as a comprehensive direct job creation program. It was a major government policy response to the problems of hardship created by a shortage of employment opportunities for unemployed workers. The program was designed “to create projects which (also) made a contribution of continuing value to the communities” (Employment and Immigration Canada, 1983b, p.2). In addition, the projects were intended to enhance the labour market experience of participants – to obtain favourable longer-term employment effects for the program clientele.

CEP had three constituent program components: Canada Community Development Projects (CCDP), Canada Community Services Projects (CCSP), and New Technology Employment Program (NTEP). The orientation of CCDP was primarily geographic and was aimed at areas experiencing relatively slow growth. For CCSP, the orientation was towards groups at a competitive disadvantage in the labour market. NTEP was designed as a wage subsidy program and targeted as a response to the problem of structural unemployment among highly qualified workers in the scientific and technology field. NTEP focused on recent labour force entrants who were unable to find employment. A stated program objective was also to provide encouragement to research and development (R&D;).

Under CCDP, the allocation was heavily concentrated in a small number of communities in the high unemployment provinces of the Atlantic region and Quebec in harmony with the program objective.

The evaluation methodology consisted of:

1. analyzing the short-term impact by conducting post-program surveys of participants and the comparison group (rejected applicants);

2. long-term labour force tracking of pre- and post-program employment experience for up to three years after program participation; and

3. analyzing the value of output of the program based on in-depth case studies involving site visits to a limited sample of projects.

The evaluation findings were as follows:

• Based on some in-depth case studies of actual projects, the evaluation concluded there was “no evidence of widespread incrementality problems with CCDP and CCSP” (EIC, 1983b, p.11). Both of these program elements provided significant employment for individuals who would not otherwise have been employed in the absence of the program -- with an average incremental employment of about 63 percent (and a range of 55 to 70 percent).

• While NTEP also produced a positive impact on employment, the incrementality was substantially small at a maximum of 29 percent. In a large number projects, the expenditure on R&D; would have been undertaken by firms without the program assistance but at a later date. Thus, NTEP brought forward the scheduling of the projects that incorporated technological change. In addition, in the converse case of employee participants, the majority of them were drawn from groups for which the probability of employment without the program was assessed to be high.

• In the short-term, CCDP participants did not show any significant increase in the proportion of time employed in the post-participation period, when compared with the experience of the comparison group. Also, there was no apparent income gain for participants after the project ended. In the case of CCSP, however, special training was positively related to the post-program employability and employment duration.

• In the longer-run, the evaluation of CCDP/CCSP showed that the average post-program employment duration, as well as the average unemployment duration, increased relative to pre-program experience.

• Around two-thirds of projects under CCDP/CCSP satisfied the standard that work resulting from these employment programs should be of continuing genuine value to the community (EIC, 1983b). The findings for NTEP were less favourable.

• A benefit-cost model was developed to assess the overall impact of the program on the economy involving wages paid and the “social opportunity cost of labour”. Evaluation studies showed that the net benefits to the economy were modest. For CCDP/CCSP, the benefit-cost ratio was 1.2. For NTEP, it was somewhat less.

The following are some major evaluation lessons that emerge from the findings:

• There are no strong arguments in favour of justifying these job creation measures in terms of their direct contribution to the economy resulting from the divergence between the social opportunity cost of labour and wage cost. The rationale for job creation programs must lie with the indirect economic benefits resulting from a more equitable allocation of job opportunities among the regions. Special training measures for the disadvantaged should be an area of important potential. In this social objective context, a program may be justified even if the cost to the economy exceeds the benefits.

• The developmental element of the program should be much more strongly linked with long-run regional development goals and modified to reflect the need for an orientation towards improving long-run employability.

• In view of the unfavourable findings for the NTEP, it was pointed out that the more appropriate program response would likely be in the form of government loans or loan guarantees rather than employment subsidies.

• There was a high incidence of recycling between program participation and UI, and therefore the linkage between the two is a matter of some concern in the longer run. Program design should reduce this type of recycling.

• Under NTEP, job creation and the encouragement of R&D; were incompatible program goals. The impact on technology development was negligible due to enrolment of inexperienced workers. R&D; did not lead to innovations but to adaptations as evidenced by the fact that they could not be patented.

The evaluation studies emphasized two crucial evaluation lessons arising out of community employment programs. Both of these pertain to faulty program design. First, the outcomes were not measurable because the likely effects were too small to be detectable under any plausible circumstances, given the uncontrolled variables in the environment of the programs. The question arose whether the programs could be more effective by offering intensive assistance to fewer target communities (EIC, 1990, p. vii). Second, a great degree of variability was built into the programs. The content of the programs had largely been for communities to decide. This also “made the outcome measurement and attribution virtually impossible “ (Watson, 1995).

CCDP was incorporated into the Local Employment Assistance and Development (LEAD) in 1983.

4.6 Local Employment Assistance and Development (1983-86)

Initiated in 1983, Local Employment Assistance and Development (LEAD) began with the objective of increasing “the number of permanent jobs in localities of chronically high unemployment”(EIC, 1986, p.62). LEAD was to fund infrastructure projects, enterprise projects, or LEAD corporations, depending on community needs and plans.

Organizations representing local development interests in communities with populations of 50,000 or less were eligible sponsors. Some 18 percent of jobs created, as reported by the LEAD corporations, were attributed to technical assistance and the remainder were attributed to financial assistance. Over 40 percent of jobs created were in manufacturing and 32 percent in services, and only 3 percent in construction. This result was a distinct difference from LIP, LEAP, and Canada Works, where a vast number of jobs were created in the construction sector.

An evaluation of LEAD was completed for 1985-86. The evaluation reported that results generally were positive and encouraging, noting in particular the low cost of jobs created⁷. The gross cost of a person-year was only $1,900 in 1986 dollars. Based on a survey of the LEAD corporations, the incrementality ratio was very low at 45 percent. This worked out to an incremental cost of $4,200. One reason why the cost is so low is “the leverage factor generated by the funds that corporations loaned/invested in their communities” (EIC, 1986, p.95). For each dollar the LEAD corporations contributed, an additional $3.80 came from other involved economic groups. Therefore, the LEAD corporations were able to create a job-creation process that increased over time.

LEAD was incorporated into the Community Futures and Job Development components of the Canadian Jobs Strategy (CJS) in 1986.

4.7 Young Canada Works (1977-80)

Introduced in 1977, Young Canada Works (YCW) had as its main objective creating short-term employment for students in areas of high unemployment. The secondary objectives consisted of: (i) creating employment that would be of some benefit to the community; and (ii) offering participants a learning and meaningful experience that they could relate to their educational goals or future career aspirations. Projects could be funded for up to 14 weeks between May and September.

In 1977, about 30,000 jobs were created under the program with an overall program expenditure of $44 million (EIC, 1978). The average number of weeks worked per job was 10.5 weeks. The gross cost per job can be estimated from the data to be $7,260 per work-year. On the whole, the majority of program participants found their experience to be directly relevant to their educational goals and career aspirations.

Regarding the distribution of funds by sponsor, the largest proportion (27.4 percent) went to Recreation and Leisure Organizations and Youth Oriented Groups. This was followed by Local and Municipal Agencies (19 percent), Service Groups and Welfare Organizations (12.5 percent), and Business and Labour Organizations plus the Private Sector (11.8 percent). Reflecting the distribution by sponsor/sponsoring organization, the following was the distribution of funds by type of activity: sports and recreation (22.5 percent), environment improvement (18.6 percent), social services (12.6 percent), research and analysis (10.7 percent), and others (35.6 percent).

The title Young Canada Works was discontinued in 1980, but similar programming was continued under Summer Youth Employment Program (1980-83). The latter program was succeeded by Summer Canada (1983- 85), and Summer Canada was brought under the Canadian Jobs Strategy in 1985⁸.

4.8 Employment Tax Credit Program (1978-81)

Employment Tax Credit Program (ETCP) was introduced in 1978 “to stimulate incremental employment in the private sector “by providing general tax incentives⁹. A secondary objective of the program was to “improve the future employability of participants above that which would have occurred in its absence”. The program offered a tax credit of up to $2.00 per hour to a maximum of 40 hours per week for a maximum of 12 months for each “eligible” worker hired (e.g. unemployed and registered with a Canada Employment Centre).

By the end of the program’s second year, some 26,000 employers participated in the program and 66,000 jobs were created, the highest participation being among companies in Quebec and the Atlantic provinces. The average duration of the subsidized job was 24 weeks

The following were some of the findings from an evaluation covering a two-year period:

• The incrementality was rather low at 37 percent. This means that 63 percent of the jobs would have been created without the program.

• The cost to government per incremental work-year, in terms of loss of revenue, was $13,500 in 1980 dollars.

• ETCP participants were more likely to draw UI benefits following termination of the subsidized employment, compared with regular workers leaving jobs about the same time.

4.9 New Employment Expansion and Development (1982-83)

New Employment Expansion and Development (NEED) program was introduced as a direct job creation initiative in 1982. It formed part of the federal government’s response to the economic recession of 1981-82. The main target group was the longer-term unemployed who had exhausted their unemployment insurance benefits and/or who were on social assistance.

NEED involved a tri-level partnership in job creation – federal, provincial and municipal governments, and the private sector. This approach marked a new era in direct job creation policy because it involved joint funding of program expenditures with the provincial government and the private sector.

Projects were required to demonstrate incremental employment growth, and sponsors were required to contribute 50 percent towards the total project cost. Project output was to be of economic value. Also, local Canada Employment Centres were to play a key role in hiring workers.

Project participants were as follows: 80 percent were UI exhaustees; 12 percent were on social assistance; and the remaining 8 percent were other unemployed persons. About 50 percent of employment was in construction, 10 percent in forest renewal, and 8 percent in high technology. During 1982-83, the federal contribution to the program was $464 Million. Provincial governments contributed about one-third of the federal contribution. In addition, the private sector contributed almost a matching amount to the federal contribution.

The evaluation findings indicated strong positive outcomes from the program. During 1982-83, the program created 88,000 jobs with an average employment duration of 23 weeks. The gross cost of job creation to the federal government (i.e. excluding cost to other partners) was estimated to be $11,900. According to selected project surveys, the average incrementality ratio was 74 percent.

4.10 Developmental Uses of UI (1977)

The Unemployment Insurance (UI) Act was amended in 1977 to design programs that would enable claimants to make constructive use of their time while on UI claim. Three developmental uses of UI funds embodied in the UI Act were for occupational training, work sharing, and direct job creation. Activities under these programs were mainly community-based, applicable to new labour market entrants as well as those who had already established UI eligibility, and generated subsequent UI entitlement. The Canada Community Development Projects and the earlier Canada Works program typified this approach. Developmental Uses of UI funds for direct job creation (permitted by Section 38 of the UI Act) was limited to UI recipients, and participation in these projects did not establish further UI eligibility after the project terminated.

The program was concentrated in three industrial sectors – forestry, mining, and fishing – which were hit hardest by the recession of 1981-82. Most projects and participants were in British Columbia and Ontario. Most of the participants were laid-off workers with UI eligibility.

The program cost mainly consisted of top-up UI benefits for project participants (estimated to be on average $2,000 more than what they would have received as regular benefits). In addition, some small expenses were granted for equipment rental or purchase.

The average duration of program participation was 13 weeks, which was short relative to other programs. The incrementality was quite high and was estimated on the basis of using a control group consisting of those who had UI eligibility but were not program participants. The incrementality was factored into the computation of cost. It is estimated (by the present authors) that the incremental cost of job creation under this program was a modest $8,500. Two factors contributed to the low cost: (a) the low opportunity cost to government of providing employment to those who otherwise would have been on UI; and (b) participation in the program did not establish further eligibility for UI following project termination.

4.11 On-Site Program (1983)

The On-Site program was developed in response to two main concerns: (a) a growing number of well educated professionals were unable to find employment in the resource management sector; and (b) a large number of organizations required temporary employees to meet their environmental responsibilities (for details see HRDC, December 1995).

The program, funded in part by Section 25 of the UI Act, provides qualified, unemployed professionals with hands-on experience addressing industrial, institutional, and municipal resource management problems. For funding to occur, participation criteria must be met by host employers as well as program workers. Employers were required to demonstrate their willingness and ability to offer incremental term employment in one of the designated program categories (which included environment management, hazardous waste management, occupational health and safety, and energy). Project workers must be in receipt of UI and possess the relevant education or work experience. The government provided enhanced UI benefits to program workers.

The following are some of the major evaluation findings and lessons learned:

• The primary conclusion from the evaluation was that the On-Site program was meeting its objectives in an effective manner and that the program was working well.

• On-Site cost of job creation per work-year is estimated to be $7,540 (estimated by the present authors from data)¹⁰. The average duration of participation in the program was 19.5 weeks. Based on a survey of employers, the incremental ratio was 76 percent (i.e. 24 percent said that they would have carried out the project irrespective of the program assistance).

• Over one-third (36 percent) of participants were hired by their On-Site employer. Of these participants, 32 percent were hired to full-time permanent positions and 50 percent were hired on contract.

• The key barrier to participation in the On-Site program in the case of workers was identified to be a lack of awareness among potential participants. Non-participants who took part in the fieldwork commented that they were unlikely to visit a Canada Employment Centre (CEC) as part of their job search because they did not believe that CECs had suitable programs or job postings for job-seekers with their background. As a result, UI recipients who did not visit a CEC or who did not know someone involved in the program were unlikely to find out about the program.

Table 5 presents a summary of the estimates of gross costs of direct job creation programs in current dollars and the incrementality ratios. Table 6 shows net or incremental costs in current and constant dollars (see also Figure 6). The results show that the cost-effectiveness of direct job creation programs has improved over time.

As can be seen in Table 5, there were wide variations in expenditure and the number of jobs created under each of the programs. The average job duration under each of the programs also varied widely from a relatively long duration of 56 weeks under Canada Community Services Projects (CCSP) and 51 weeks under Local Employment Assistance and Development (LEAD) , for instance, to a relatively short duration of 10.5 weeks under Young Canada Works.

The average incrementality also varied widely by program. The programs which had relatively high incrementality ( 75 percent or over ) were LIP, Canada Works, and On-Site. Those with relatively low levels of incrementality ( 50 percent or less) were NTEP, LEAD, and Employment Tax Credit Program.

Table 6 shows the incremental cost per job in constant dollars. It can be easily seen that with a significant shift from passive to active use of Unemployment Insurance program under Developmental Uses of UI, job creation programs have become significantly more cost effective. Compare, for instance, the incremental cost of job creation of $9,600 under developmental uses of UI with over $20,150 under the Employment Tax Credit Program, or $17,000 under Canada Works ( General). It is the continuous process of evaluation and assessment of programs, and modifying them to enhance the elements that worked that contributed to the cost-effectiveness.

View Table 5

¹ No precise estimate of incrementality was possible for Young Canada Works. Based on data for similar programs, an incrementality of 60 percent has been assumed .

View Figure 6

5.1 Evaluation Lessons

The following evaluation lessons can be drawn from the analysis presented in this review.

Canadian direct job creation programs have generally served as useful counter-cyclical policy tools by creating temporary jobs quickly and on a short notice.

Programs, such as LIP, Canada Works, LEAP, and Canada Employment Program, were created on a short notice in times of economic downturn and were discontinued as the economy improved. These programs were administered through municipalities and non-profit community organizations without large expenditures on creating administrative infrastructure.

Most of these programs were assessed to have funded projects that were considered to be of value to the community.

Although most of the programs were found to have funded valuable projects, several evaluation studies found that measurable indicators for “value to community” were not defined. Therefore, it was often difficult to evaluate the program outcome from the point of community value. The result suggests that it would be useful to define more precisely some measurable indicators of value added for the community.

Direct job creation programs were also generally found to be appropriate tools for targeting certain disadvantaged labour force groups.

Programs, such as LIP, Canada Works, LEAP, and Canada Community Services Projects, addressed the special needs of the unemployed and social assistance recipients. The New Employment and Expansion Development (NEED) program targeted the long-term unemployed (UI exhaustees). Evaluation findings generally confirmed that participant profile conformed to the intended target groups. It may be pointed out, however, that the direct job creation programs did not specifically target women, and the participation level for women was often lower than their share in the overall unemployed.

The use of job creation programs for addressing concerns relating to regional structural unemployment was not always successful.

Tension between conflicting goals of a given program has often been highlighted in evaluations.

Over time, the cost effectiveness of the direct job creation program has improved considerably.

Although employment incrementality still remains at a modest level, costs of job creation have become lower over time due to developing better program design, dropping ineffective elements, and expanding better elements. Community-based delivery has generally worked well. In recent years, job creation within the framework of Developmental Uses of UI Funds has used UI entitlement rather than simply helping workers re-qualify for UI.

Targeted direct job creation has worked better than the use of general tax concessions.

For instance, the Employment Tax Credit Program (ETCP) had a higher cost and much lower incrementality ratios than several wage-based job development programs. In addition, ETCP participants had higher probability of going on UI after subsidized employment, compared to regular workers.

Combining the objectives of job creation and promoting scientific and technical research within the framework of the same program does not work.

The New Technology and Employment Program (NTEP) component of the Canada Employment Program made an attempt to combine these two goals under the same program and they were incompatible. In order to be able to have the maximum impact on job creation at a relatively low cost, it is appropriate to target the unemployed and inexperienced work force. These groups are, however, unable to contribute significantly towards scientific research.

Job creation incrementality is low in the private sector relative to the public and non-profit sectors.

Public works can mobilize temporary jobs quickly in times of economic downturn and they can be limited to slow growth areas during economic upturns. Federal partnership with provincial and municipal governments in job creation, as in New Employment and Expansion Development (NEED), has been a success story.

The role of direct job creation programs in developing skills and improving the employability of program participants after the program ends remains an open question.

In many instances, program participants were found to have gone back on UI or social assistance after the program ended. Developmental Uses of UI Funds has largely plugged this loophole in recent years.

More consistent and new evaluation approaches are needed to assess the net impacts of job creation programs.

In program evaluations, estimates of incrementality have been usually based on participant employer and employee surveys. Needless to say, these are subjective and less reliable than other methodologies. Where possible, macro-economic analysis should be considered as a way to establish the impact of job creation programs on the total number of jobs. In future evaluations of job creation programs, it is imperative that a standard outcomes approach be adopted that considers four accountability measures: (a) net cost per net job created; (b) net cost per person off the unemployment count; (c) reduction in the unemployment count as a percent of persons supported; and (d) net jobs created as a percent of those supported.

Planning the creation of an appropriate data set appears to be a key element in reliable program evaluation.

In many instances, lack of management and appropriate data collection has hampered proper and timely evaluation. It seems imperative that, before a program is mounted, decisions are made about evaluation issues that will be addressed. At the same time, data requirements need to be specified and appropriate arrangements need to be put in place to meet the data requirements. In recent years, this deficiency has been met (e.g. SEA and other recent indirect job creation programs).

Using Canada Employment Centres (now called Human Resource Centres of Canada) is a cost effective delivery mechanism.

Although using Canada Employment Centres (CECs) is cost effective, several evaluation studies have pointed out that non-participants in several programs were not aware of the availability of the programs because the non-participants were not enrolled with a CEC. It would, therefore, be important to diversify the delivery mechanism to reach the clientele (especially semi-professional and professional groups) who might not enrol themselves as unemployed in a CEC/HRCC (Human Resource Centre of Canada).

The Canadian experience with direct job creation programs generally falls in line with that of most other OECD countries. Similarities and differences are briefly highlighted below.

6.1 Other OECD Countries

In other OECD countries, the popularity of direct job creation programs increased considerably in the mid-seventies with rising unemployment. Similar to Canadian job creation programs, the target groups have usually been the general unemployed, long-term unemployed, older workers, young people, women, and new labour force entrants¹¹. Also, the eligible sponsors were similar to those in Canadian programs: local authorities, voluntary and charitable organizations, and sometimes also private sector employers. Almost all programs specified “additional” employment as a requirement for eligibility.

Major program funding was generally provided by the federal governments and in the form of subsidised wages of participants (usually ranging from 60 to 100 percent). In Denmark, however, the subsidy was 120 percent of gross wage costs. In the U.S. and U.K., some projects require supplementary program funding from other levels of government.

As in Canada, projects in other OECD countries generally occurred in the following sectors: construction, urban renewal, forestry, and environmental areas. The project duration did not usually exceed 52 weeks (18 months in the U.S.).

6.2 The United States

There are two major direct job creation programs in the U.S., and these are funded under Comprehensive Employment and Training Act (CETA). One program is designed for the disadvantaged who have been unemployed for 15 weeks or more (Title II. D). The other is for those who have been unemployed for 10 weeks or more and who come from low-income families (Title VI). The intent is that Title II.D should have basically a structural impact, while Title VI is expected to serve counter-cyclical purposes.

With the use of two programs, two objectives of job creation are kept somewhat distinct. In most Canadian job creation programs, however, the distinction is not so clear-cut. Whether this is desirable or not is discussed below.

6.3 Two Key Issues

In the light of a brief review of direct job creation programs in other OECD countries, two issues appear to be of major importance.

The first issue concerns the job displacement effect and the related level of incrementality in job creation. As in the case of program evaluation in Canada, other countries derive the incrementality estimates from sample surveys. These surveys are necessarily subjective and hence not strictly comparable. While recognizing the limitations of these estimates, however, the incrementality ratios of Canadian job creation programs compare quite favourably with those reported in other countries (50 to 60 percent in the U.K., and 65 to 70 percent in the U.S.; see OECD, 1980). Due to more targeted programming in recent years, job creation programs in Canada seem to have reduced job displacement and achieved a larger measure of program incrementality.

The second issue relates to the objectives of direct job creation programs. As temporary employment measures, they have two broad objectives: (a) to provide counter-cyclical economic stimulus; and (b) to reduce structural problems among certain demographic groups, regions, or industry sectors. Ideally, as a counter-cyclical measure, job creation programs should be capable of a fast phasing-in and phasing-out, be highly labour intensive, have low displacement, and draw clientele from the unemployed but not from outside the labour force. To address structural labour market problems, however, the speed of phasing-in and phasing-out is not so important. In fact, to phase-in a series of projects designed to give maximum adjustment assistance to groups with structural labour market problems probably takes time, and the phasing-out of successful projects may not even be desirable (OECD, 1980). Also, there is no rationale for restricting such programs to non-profit activities. In many instances, therefore, counter-cyclical and structural goals may be inconsistent objectives within the framework of the same program.

The analysis presented in this report indicates that tension between competing objectives is embedded in several Canadian programs. In several other countries, especially in the U.K. and U.S., the counter-cyclical and structural elements have been kept distinct from one another and not combined within the fold of the same job creation program. This important consideration should be kept in mind in future program design in Canada.

Balkenhol, Bern, 1970. “Direct Job Creation in Industrialized Countries”, International Labour Review, July-August, 120:4.

Canada House of Commons, 1981. Work for Tomorrow: Employment Opportunities for the ‘80s, Ottawa. (Also popularly referred to as the Allmand Report).

Cullen, Bud. Shape and Scale of the Employment Strategy for 1978-79, Discussion paper, June.

Danzinger, Sheldon and Daniel H. Weinberg ( Eds.) 1986. Fighting Poverty, Harvard University Press, Cambridge.

Davis, Steven J., 1996, Job Creation and Destruction, MIT Press.

Employment and Immigration Canada Annual Reports, various years.

Employment and Immigration Canada, 1978, Evaluation of the 1977 Young Canada Works Program, Program Evaluation Branch, February.

Employment and Immigration Canada, Evaluation of the Economic Growth Component of Canada Works: Final Report, Program Evaluation Branch, April.

Employment and Immigration Canada, 1981a, Labour Market Development in the 1980s (Also popularly referred to as the Dodge Report).

Employment and Immigration Canada 1981b, Evaluation of Three Job Creation Programs, Volume I, Methodology, Program Evaluation Branch, July.

Employment and Immigration Canada 1981c, Youth Job Corps Evaluation Report, Program Evaluation Branch, August.

Employment and Immigration Canada 1982a, Evaluation of the Employment Tax Credit, January.

Employment and Immigration Canada 1982b, An Evaluation Assessment of the Local Employment Assistance Program, March.

Employment and Immigration Canada 1982c, Evaluation of Three Job Creation Programs, Volume II, Program Implementation Review, Program Evaluation Branch, June.

Employment and Immigration Canada 1982d, Evaluation of Three Job Creation Programs, Volume III, Value of Output of CCDP and CCSP, Program Evaluation Branch, June.

Employment and Immigration Canada 1982e, Evaluation of Three Job Creation Programs, Volume IV, Value of Output and Labour Market Impact of NTEP, Program Evaluation Branch, September.

Employment and Immigration Canada 1982f, Local Economic Development Assistance Program, Program Evaluation Branch.

Employment and Immigration Canada 1982g, Evaluation of the Employment Tax Credit Program, Program Evaluation Branch, January.

Employment and Immigration Canada 1983a, Evaluation of Three Job Creation Programs, Volume V, Analysis of Longitudinal File Data and Survey, Program Evaluation Branch, March.

Employment and Immigration Canada 1983b, Evaluation of Three Job Creation Programs: Summary and Review of Findings and Policy Recommendations, Program Evaluation Branch, August.

Employment and Immigration Canada 1984, Evaluation—Developmental Use of UI Funds For Job Creation, Final Report, Prepared by Abt Associates for Program Evaluation Branch, July.

Employment and Immigration Canada, The Evaluation of the Local Employment Assistance and Development (LEAD), by L. Ladouceur and P. Kinoshameg, Program Evaluation Branch, November.

Employment and Immigration Canada 1990, Evaluation of Community Futures Program, Program Evaluation Branch.

Faulkner, M.J. 1978, Direct Job Creation Schemes in the Member States of the European Community.

Fechter, A., 1975, Public Employment Programs, Evaluative Studies No.20, American Enterprise Institute for Public Policy Research.

Guest, Carol, Fred Phelan, and Zhenxi Lin. 1993, Unemployment Insurance: Evaluation Lessons Learned, Human Resources Development Canada, Program Evaluation Branch, October.

Hawkes,F.J., W.D. Gabbert, and V.M. Bryant, 1973. The Local Initiatives Program Evaluation, School of Social Welfare, University of Calgary.

Human Resources Development Canada, 1995. Evaluation of the On-Site Program, December.

Human Resources Development Canada, 1997. Lessons Learned: Effectiveness of Employment-Related Programs for Youth, Evaluation and Data Development, June.

Hunter, John, The Employment Challenge, Employment and Immigration Canada, Undated, Ottawa.

Jackson, M.P. and Hanby, V.J.B. (Eds.), 1979. Work Creation-International Experiences, Saxon House.

Manpower and Immigration,1981. Evaluation of the Community Employment Strategy.

OECD, 1980. Direct Job Creation in the Public Sector, Paris.

OECD, 1997. Employment Outlook, July, Paris.

OECD, 1998. Employment Outlook, July, Paris.

Roy, Arun S., 1984. Employment Impacts of Alternative Government Initiatives and Related Evaluation Criteria, Employment and Immigration Canada.

Watson, Kenneth, 1995. “Program Design Can Make Outcome Evaluation Impossible : A Review of Four Studies of Community Economic Development Programs”, The Canadian Journal of Program Evaluation, Vol. 10, No. 1.

Footnotes

1	This approach is particularly relevant in assessing the cost-effectiveness and success of training programs. For example, it uses labour market performance indicators such as the pre- and post-program probability of finding employment. Also, it can compare pre- and post-program earnings, employment duration, incidence of UI and welfare, etc.
2	See Employment and Immigration Canada, 1982 c, pp.4-5.
3	For a detailed discussion of evaluation criteria and estimates of costs to the federal government of job creation under alternative fiscal initiatives see Roy (1984). The study uses a macro-modelling methodology to estimate the net cost of job creation to the federal government in a national accounting framework under general tax and expenditure initiatives as well as more specific expenditure initiatives directed toward particular industry sectors. The study takes into account the primary as well as secondary impacts of federal programming.
4	On direct job creation programs, the Dodge report concluded that direct short-term job creation programs that could fund local projects had a significant benefit in periods of cyclical downturn. Projects could be mounted quickly, accurately targeted and easily terminated. The Allmand report agreed on the usefulness of direct job creation projects but recommended that all job creation projects include a training element.
5	EIC Annual Report, 1985-86, p.18.
6	Derived from a survey of managers, 1972-73 projects (Hawkes et. al., 1973).
7	L. Ladouceur and P. Kinoshameg, The Evaluation of the Local Employment Assistance and Development Program, Ottawa, EIC, November 1986.
8	For a detailed review of youth initiatives see Human Resources Development Canada, Lessons Learned: Effectiveness of Employment-Related Programs for Youth, Evaluation and Data Development, June 1997.
9	For details see EIC, January 1982a.
10	Includes only the cost to government in terms of enhanced UI benefit payments to program workers; does not include costs to the employers (e.g. purchase of project-specific equipment, etc.)
11	In the U.S., two principal direct job creation programs are covered under the Comprehensive Employment and Training Act (CETA): one targeted to the general unemployed; and the other targeted to low-income and disadvantaged groups. In the U.K., the Job Creation Program (JCP) was introduced in 1975 and was conceived much along the lines of the Canadian LIP. JCP has recently been replaced by two separate programs: Youth Opportunities Scheme (restricted to unemployed young people); and Special Temporary Employment Programme (targeted to older people with long-term unemployment). In Denmark the major job creation programs are Public Employment Works (PEW) and Employment Projects for Young People (EPY); the two programs target all unemployed and young workers respectively. Programs in Norway, called Extra Employment Scheme (EES) and Alternative Jobs in the Public Sector (AJPS), have a similar orientation as in Denmark. For details, see OECD, 1980; Jackson and Hanby,1979; and Davis, 1996).