• Governments in Canada do little, compared to those in the United States, to support the re-entry of people with criminal records into the workplace.
• Employment is a key part of re-entry. There are significant positive externalities associated with employment, including less recidivism and safer communities. Yet employers tend to overestimate the risk associated with hiring people with criminal records and are often unwilling to accept this perceived risk. This justifies government action to incentivize them to hire more from this population.
• Governments at all levels should implement a made-in-Canada version of the US Federal Bonding Program. Such a program shifts the perceived risk of hiring these workers from employers to governments. It provides insurance to employers, at no cost to them, against losses from theft or fraud if committed by members of these groups.
• Work bonds can be effective in encouraging employers to hire people with criminal records, and the US experience shows that the program is highly cost-effective, supporting the employment of most participants at a cost to the taxpayer of a mere US $100 each.
• This paper outlines a detailed implementation strategy for a Canadian version of the Federal Bonding Program that would cost $800,000 to support the employment of nearly 6,000 people. A successful program will be simple to navigate, administratively lean, and very affordable for the taxpayer.

Work is fundamental to human existence. Of course, it is essential to our economy and to producing the goods and services that humanity needs to survive. But work is not just a means to that end. It is an inherent part of human flourishing. Work engages our creativity, our talents, our bodies, and our minds. Through work, we create something of value and put it at the service of the community. Work is one of the aspects of our lives that provides meaning and direction.

As such, employment provides significant dividends to workers aside from the benefit of a paycheque. People with jobs have better physical and mental health. They are more likely to have stable family lives and less likely to engage in self-harm. And they are less likely to be involved in crime. 1 1 Evidence for these statements is provided in B. Dijkema and M. Gunderson, “Work Is About More Than Money: Toward a Full Accounting of the Individual, Social, and Public Costs of Unemployment, and the Benefits of Work,” Cardus, 2019, https://cardus.ca/research/work-economics/reports/work-is-about-more-than-money/; R. Nauta, “Curbing Crime with Employment: Exploring Work as Crime Prevention for Canadians with Criminal Records,” Cardus, 2023, https://cardus.ca/research/work-economics/reports/curbing-crime-with-employment/.

Yet many people have difficulty getting a job because of challenges in their lives or mistakes they have made, such as those who have criminal records or who are recovering from addiction. Their difficulty finding employment may be due in part to employers’ reluctance to hire them, which in turn may be due to inaccurate assumptions about them. For people like these, governments can play a role in breaking down the employment barriers that they face.

This paper recommends a new policy for governments in Canada to encourage employers to hire people who have criminal records, are in recovery from addiction, or have other characteristics that make finding a job challenging. It argues that governments in Canada—whether federal, provincial, or even municipal—should introduce a made-in-Canada version of the US Federal Bonding Program.

Cardus initially suggested this policy in a paper co-published in 2023 with the University of Toronto’s Munk School of Global Affairs and Public Policy. 2 2 R. Nauta, “Employing Ontarians with Criminal Records: Policies to Correct Labour Market Misconceptions,” Ontario 360, 2023, https://on360.ca/policy-papers/employing-ontarians-with-criminal-records-policies-to-correct-labour-market-misconceptions/. The current paper incorporates new insights that have arisen from more research and includes a detailed policy design and implementation strategy.

The current paper also seeks to broaden the policy recommendation in two ways: jurisdictionally, and in terms of the population covered.

First, because the earlier paper was written for the Munk School’s Ontario 360 project, it focused on the Government of Ontario only. Given that the bonding program could be applied to any jurisdiction, however, this paper analyzes the policy in a way useful to policy-makers at any level of government and across Canada.

Second, the earlier paper focused exclusively on people with criminal records. A bonding program such as the one proposed here, however, could cover any group of people whom employers may be reluctant to hire due to concern about potential losses from theft. That said, people with criminal records make up the largest group of users of the Federal Bonding Program, so that will continue to be the focus of this paper.

This paper proceeds as follows. The first section examines the relationship between employment and criminality. The second section examines the US program and considers its effectiveness and limitations. The final section outlines the policy considerations and costing for a Canadian version of this policy.

Having a job offers a path forward for people with criminal records to set their lives back on a legal and pro-social foundation. At the same time, having a criminal record makes it difficult for people to find a job, producing a Catch-22 situation. 3 3 T. G. Chiricos, “Rates of Crime and Unemployment: An Analysis of Aggregate Research Evidence,” Social Problems 34, no. 2 (1987):187–212, https://doi.org/10.2307/800715; D. B. Mustard, “How Do Labor Markets Affect Crime? New Evidence on an Old Puzzle,” IZA Discussion Paper no. 4856 (2010), https://docs.iza.org/dp4856.pdf; Nauta, “Curbing Crime with Employment”; D. Pager, “The Mark of a Criminal Record,” American Journal of Sociology 108, no. 5 (2003): 937–75, https://doi.org/10.1086/374403.

It is important to acknowledge the reasons why employers may be disinclined to hire someone with a criminal record. Employers may be concerned that the employee would be more likely to steal from them. They may want to protect other employees, customers, and clients from potential harm. They may interpret someone’s previous involvement in crime as a sign that they lack a strong work ethic. And they may believe that their business insurance requires them to check criminal records and not hire those who have them.

The Federal Bonding Program targets the first of these concerns, but all of them are understandable to some degree. There is a well-established relationship between unemployment and criminality, although the precise nature of the relationship is unclear. Some researchers posit that unemployment and criminality are each a manifestation of a deeper problem, a lack of personal self-control. 4 4 M. R. Gottfredson and T. Hirschi, A General Theory of Crime (Stanford University Press, 1990). Others argue for a causal relationship: that unemployment contributes to the likelihood of engaging in crime, or vice versa. The majority of researchers, however, believe that increasing the employment rate of people with criminal records will reduce their propensity to commit further crime, either because employment income reduces the motivation to obtain money by criminal means, or because the job itself provides important social bonds that promote law-abiding behaviour. 5 5 G. S. Becker, “Crime and Punishment: An Economic Approach,” Journal of Political Economy 76, no. 2 (1968): 169–217; R. J. Sampson and J. H. Laub, Crime in the Making: Pathways and Turning Points Through Life (Harvard University Press, 1993).

Of course, even if people with criminal records are, overall, less likely to recidivate when they have a job, the risk to any given employer is not zero. Employers’ hiring decisions involve calculation of risk. Employers do not perform a rigorous actuarial assessment of an applicant’s likelihood of recidivating but make decisions based on their perceptions of risk. The higher the perceived risk in hiring someone, the higher the anticipated cost.

A large body of evidence indicates that the actual risk is lower than employers perceive it to be. 6 6 The evidence is discussed in Nauta, “Curbing Crime with Employment.” One study has found that 15 percent of employers say they would refuse to hire someone with any criminal record at all, regardless of its contents, 7 7 John Howard Society of Ontario, “Help Wanted: Reducing Barriers for Ontario’s Youth with Police Records,” 2014, http://johnhoward.on.ca/wp-content/uploads/2014/07/johnhoward-ontario-help-wanted.pdf; John Howard Society of Ontario, “Invisible Burden: Police Records and the Barriers to Employment in Toronto,” 2018, https://johnhoward.on.ca/download/the-invisible-burden/. yet in some contexts and on certain metrics, people with criminal records perform their work as well as or better than other employees. 8 8 J. H. Lundquist, D. Pager, and E. Strader, “Does a Criminal Past Predict Worker Performance? Evidence from One of America’s Largest Employers,” Social Forces 96, no. 3 (2018): 1039–68, http://dx.doi.org/10.1093/sf/sox092; D. Minor, N. Persico, and D. M. Weiss, “Criminal Background and Job Performance,” IZA Journal of Labor Policy 7, no. 8 (2018): 2, https://doi.org/10.1186/s40173-018-0101-0. Treating every criminal record as a veto on employment is at odds with the body of research revealing that the risk is overestimated.

The preponderance of research evidence suggests that employers are not availing themselves of a labour pool that could provide value to their businesses. This in and of itself may not be enough to justify government action, since firms should generally be trusted to act as they think best. But it does indicate that policies intended to increase employment levels of people with criminal records may benefit employers as well as society as a whole.

If the majority of scholarly opinion described in the previous section is correct, the Catch-22 presents a particular challenge for policy-makers, because it means that the labour market on its own is unlikely to achieve the socially optimal level of employment for people with criminal records. The positive externalities for society at large—less recidivism, safer communities, and so on—fall outside the rational decision-making of employers. The benefit to society justifies action by government to incentivize greater employment of this population than employers would naturally offer. There are several possible interventions, but three main types are relevant to this paper’s analysis 9 9 For an alternate categorization, see R. Nauta, “Humanizing Criminal Records: Toward a View of the Whole Human Person,” Cardus, 2023, https://cardus.ca/research/work-economics/reports/humanizing-criminal-records/. :

• Improve the overall employability of people with criminal records. Measures could include increased skills training and drug rehabilitation in prisons. In theory, by making this group more employable overall, employers would be more inclined to hire from it. But since employers’ perceptions of the hiring risk generally do not match reality, this category of intervention would likely need to be coupled with others, described below.
• Change employers’ perspectives. Public awareness campaigns and promotion by political leaders generally fall into this category. Cardus has previously proposed a new framework for releasing criminal records that would provide employers with a more complete, and therefore more accurate, picture of the human person behind the record. 10 10 Nauta, “Humanizing Criminal Records.” Yet, while this category of policies may be beneficial in the long run, it may be less likely to have short-run results, because perceptions take time to change.
• Shift the risk from employers to government. In any market transaction, where there is uncertainty, someone has to bear its cost. If it is correct that society benefits when a greater number of people with criminal records have jobs, there is a case for government to bear some of this risk. And if employers have an exaggerated sense of the risk associated with hiring from this group, the amount of actual risk that government would have to take on would be smaller than the amount of perceived risk that employers would be relieved of. The government could achieve a great deal of perceived-risk reduction for a comparatively small cost. Because this policy lever would influence actions without necessarily changing perceptions, it could have a quicker effect than the two other options outlined above. And because it directly reduces the cost to employers, the effect is likely to be more predictable than other, more indirect, policies. The Federal Bonding Program falls into this category.

Since 1966, the US federal government has provided insurance to employers, in the form of a bond, for losses from theft, forgery, larceny, fraud, or embezzlement committed by employees who are members of the following groups:

• people with criminal records
• welfare recipients
• those in recovery from addictions
• those with poor credit records
• economically disadvantaged people who lack work histories
• people dishonourably discharged from the military 11 11 Unless otherwise noted, all information in this section is taken from the Federal Bonding Program website at https://bonds4jobs.com or from conversations with Tom Villanova, director of the Federal Bonding Program, on April 2 and May 27, 2024. All figures relating to the Federal Bonding Program are given in USD, unless otherwise noted.

Bond values range from $5,000 to $25,000, depending on the responsibility involved in the occupation. Most of the bonds issued are for the minimum amount of $5,000.

The government pays for the bond at a rate of about $100 per $5,000 of coverage. 12 12 A range of $84–$98 per $5,000 bond was quoted in U.S. Department of Labor, Training and Employment Notice No. 37-07 (March 27, 2008), https://dol.gov/sites/dolgov/files/ETA/advisories/TEN/2008/TEN37-07.pdf. The employer pays nothing, and there is no deductible. The bond covers any loss to the employer up to the amount of the bond; the employer is responsible for the portion of any loss that exceeds the amount.

Obtaining the bond is straightforward. The prospective employee contacts the state bonding coordinator; usually this involves a simple phone call. The state bonding coordinator fills out the necessary information to create the bond and then sends it to the prospective employer. There is no action needed on the part of the employer. The fact that there is virtually no red tape for businesses is a key feature of the program. Moreover, no applications are denied, provided that the prospective employee is a member of one of the targeted groups.

The bond covers the first six months of employment. The employee can then apply for a second six-month bond with the same employer, provided that the first bond resulted in no claim. After the twelve months, the employee is no longer eligible for a bond with the same employer but remains eligible for any future employers.

The program is administered by Union Insurance Group, a company that specializes in providing insurance to labour organizations, through a contract from the US Department of Labor. It is underwritten by Chubb, one of the world’s largest insurance companies, although under a 2019 expansion of the program discussed below, individual states are able to contract out to a different insurance provider if they so choose. 13 13 U.S. Department of Labor, Advisory: Training and Employment Guidance Letter No. 15-18 (March 26, 2019), https://dol.gov/sites/dolgov/files/ETA/grants/pdfs/TEGL-15-18.pdf.

The program administration is remarkably lean, with only two direct employees: the manager of the program and one IT support person. The rest of the administration is done in coordination with the state governments, with each designating a state bonding coordinator, whose responsibility for liaising with the program is usually in addition to other responsibilities.

Although the Federal Bonding Program has existed for over half a century, there is a notable lack of scholarly studies on the program itself, with one researcher remarking that “information on the program is surprisingly scarce.” 14 14 S. Bushway and M. Denver, “Review of The New Scarlet Letter? Negotiating the U.S. Labor Market with a Criminal Record,” ILR Review 68 no. 4 (2015): 963–65, https://jstor.org/stable/24812016. Nevertheless, there is reason to believe that the program is successful in increasing employment among people with criminal records.

The program reports a 99 percent success rate, meaning that 1 percent of the bonds issued by the program end in a claim. This rate suggests that the program either deters theft very effectively, or that it is used almost entirely by those who would not have committed this crime even absent the bond. Although they may have been low-risk job applicants, the fact that these employees sought a bond in the first place suggests that employers were reluctant to hire them without it. This in turn suggests that the program is achieving its purpose in helping deserving people to get jobs that they might not otherwise have gotten.

While there is a dearth of studies on the program itself, a handful of recent studies have tested employers’ reactions to hypothetical insurance systems that resemble the US program. These studies provide evidence that bonding programs can be effective at increasing the employment rate of the targeted groups. One such study tested the effect on employers’ hiring decisions under scenarios in which different amounts of insurance were available to them. In this study, the willingness of employers to hire people with criminal records increased from 39 percent to 51 percent when they were offered a $5,000 insurance policy. Notably, larger insurance policies with coverages of $100,000 or even $5 million did not significantly increase the willingness of employers to hire from this population. 15 15 Z. B. Cullen, W. S. Dobbie, and M. Hoffman, “Increasing the Demand for Workers with a Criminal Record,” National Bureau of Economic Research, 2022, https://nber.org/papers/w29947. In other words, a little bit of insurance goes a long way.

Another study found that while insurance programs are effective in incentivizing hiring of people with criminal records, the degree of impact on employers’ hiring decisions was dependent on the nature of a prospective employee’s offence. Notably, the experiment found that providing an insurance policy similar to the Federal Bonding Program almost completely mitigated the negative effect of a misdemeanor drug offence on an employer’s propensity to hire. 16 16 S. Bushway and J. T. Pickett, “Direct Incentives May Increase Employment of People with Criminal Records,” RAND Corporation, 2023, https://rand.org/pubs/working_papers/WRA2786-1.html. This suggests that bonding programs can also be effective in increasing employment among another of the Federal Bonding Program’s targeted groups, that is, those recovering from substance addictions, even if they do not have criminal records.

Another key question is whether the results of the program are worth the costs. On this measure, the program is probably among the most cost-effective policies on offer for increasing the employment of the targeted groups. Although the program does not publish its overall cost, a simple calculation shows that it cannot be more than a few hundred thousand dollars in insurance premiums. 17 17 As stated previously, the premiums, paid by government, are $100 per $5,000 of coverage. In a recent year, there were 1,691 bonds issued to 1,068 people, and most of them were for the lowest coverage amount. Even if half of the bonds were for the maximum coverage amount, the total outlay would be $507,000. Given that most bond recipients receive the minimum coverage amount, the government is spending a mere $100 per person to incentivize employers to hire them.

Moreover, one of the studies cited above found that a bonding program could have as much impact on incentivizing employment as a wage subsidy that covered 80 percent of an employee’s earnings. 18 18 Cullen et al., “Increasing the Demand for Workers with a Criminal Record.” For comparison purposes, a wage subsidy of that magnitude would likely cost the government $12.8 million to incentivize the employment for six months of the same number of people (1,068) as were covered by the program in a recent year. 19 19 Assuming full-time work at New York State’s minimum wage as of May 2024 ($15 per hour), the earnings of 1,068 people over six months would equal approximately $16 million. A wage subsidy covering 80 percent of that would cost the government $12.8 million.

A new or revitalized program, if properly advertised, could potentially reach a larger number of targeted workers and therefore entail a larger expenditure by government. The main point is that an insurance policy like that of the Federal Bonding Program is likely much more cost-effective than other policy options.

The program has important limitations, but these are not inherent in the program design. Some could be solved by increased awareness of the program among both prospective employees and employers, through advertising and promotion. Others are simply owing to the fact that the program does not solve all the problems of employment for the  people with criminal records. But no single policy ever could. Rather, the program must be seen as one component of a broader policy agenda.

The most obvious limitation of the program has been its very small size in terms of the number of people supported. Although almost eight million people in the United States have been imprisoned at least once in their lives, 20 20 T.-A. Craigie, A. Grawert, and C. Kimble, “Conviction, Imprisonment, and Lost Earnings: How Involvement with the Criminal Justice System Deepens Inequality,” Brennan Center for Justice, 2020, https://brennancenter.org/sites/default/files/2020-09/EconomicImpactReport_pdf.pdf. the program has provided bonds for only 56,500 people since its creation in 1966. 21 21 “About the FBP,” Federal Bonding Program, https://bonds4jobs.com/about-us. Similarly, although about 600,000 are released from prison every year in the US, 22 22 “Incarceration and Reentry,” U.S. Department of Health and Human Services, Office of the Assistant Secretary for Planning and Evaluation, https://aspe.hhs.gov/topics/human-services/incarceration-reentry-0. the program issued bonds to only 1,068 people in a recent year. 23 23 “About the FBP,” Federal Bonding Program. By either measure, less than 1 percent of formerly incarcerated people have participated in the program. In fact, the percentage must be even lower, since the program also covers members of other target groups. 24 24 Although the program does not publish the distribution of bonds across the different target groups, the program director confirmed that approximately three quarters of the bond recipients are justice-involved. Conversation with Tom Villanova, April 2, 2024.

In 2019, the US government provided an additional $5 million in grants to the states to expand the use of bonds structured like those in the federal program. 25 25 U.S. Department of Labor, Advisory: Training and Employment Guidance Letter No. 15-18. That year, twenty-four states signed up for the expanded program, most of them receiving a grant of $100,000 each. 26 26 U.S. Department of Labor, U.S. Department of Labor Awards $2,234,122 to 24 States for Fidelity Bonding Demonstration Grants (June 13, 2019), https://dol.gov/newsroom/releases/eta/eta20190613-1. In 2020, the government re-opened applications to the fund to encourage additional states to participate, which led an additional seven and the District of Columbia to do so the following year. 27 27 U.S. Department of Labor, Advisory: Training and Employment Guidance Letter No. 10-20 (December 21, 2020), https://dol.gov/sites/dolgov/files/ETA/advisories/TEGL/2020/TEGL_10-20.pdf; U.S. Department of Labor, U.S. Department of Labor Awards $725K to Help At-Risk Workers Overcome Barriers to Employment (April 7, 2021), https://dol.gov/newsroom/releases/eta/eta20210407. While it is a hopeful sign that the federal government added money to the program and that thirty-two jurisdictions signed up for it, it is unclear why the remainder elected not to.

So why the low uptake within the target population? One study suggests that it is due to “difficulty navigating the program requirements” or a lack of knowledge about the program’s existence. 28 28 Cullen et al., “Increasing the Demand for Workers with a Criminal Record.” Since the program’s administrative burden is in fact quite light, the first point speaks to the importance of clearly communicating to prospective employers and employees how simple the application process actually is.

That leaves lack of knowledge as the most likely reason for the continued low use. The director of the program has noted that having the US Chamber of Commerce post a link to the program on its website may have contributed to greater uptake. 29 29 Conversation with Tom Villanova, June 18, 2024. The grants provided in 2019 and 2021 allowed each state to spend up to 15 percent of its allocation on advertising the program. For most states, this represents a mere $15,000, which is not likely to go far, especially in states with large populations. It seems that there is scope for expanding the advertising of the program or finding alternate ways to promote it.

Another limitation of the program is that while it can successfully change employers’ decisions and increase employment of people with criminal records, it is unlikely on its own to change employers’ perceptions of this population. One experiment showed that hiring incentives such as insurance policies and wage subsidies could increase employment but had no effect, positive or negative, on employers’ perceptions of those with a criminal record. 30 30 Bushway and Pickett, “Direct Incentives May Increase Employment of People with Criminal Records.” This underscores the point that a bonding program alone will not be sufficient to overcome all the employment challenges that the target population faces. It will have to be combined with other efforts that seek to correct the perspectives of employers, which will, in turn, have the effect of catalyzing employers’ interest in the program. 31 31 Although it is beyond the scope of this paper, such policy options are the topic of Nauta, “Humanizing Criminal Records.” That paper proposed a new conceptualization of criminal records aimed at motivating employers to take a whole-person perspective in hiring people with criminal records. That paper also included a list of policy options from existing academic research.

Some research has suggested that subsidies can have the unintended consequence of reinforcing negative perceptions against the groups they aim to help and thus undercutting their chances in the job market. 32 32 L. F. Katz, “Wage Subsidies for the Disadvantaged,” National Bureau of Economic Research, 1996, http://nber.org/papers/w5679.pdf. Former US Assistant Secretary of Labor Brent Orrell has suggested that the program can act as a “scarlet letter” for people with criminal records, which led him to consider alternative solutions for increasing the employment rate of this group. 33 33 White House Faith-Based and Community Initiatives, Compassion in Action: Improving Prisoner Re-Entry Services Through Faith and Community-Based Partnerships (March 22, 2007), https://georgewbush-whitehouse.archives.gov/government/fbci/pdf/march_cia_roundtable_report.pdf; U.S. House of Representatives, An Examination of Federal Employment Practices/Policies in Hiring Ex-Offenders (June 10, 2008), https://govinfo.gov/content/pkg/CHRG-110hhrg48239/pdf/CHRG-110hhrg48239.pdf. Rather than undermining the argument for the program, this objection should be understood as further underscoring the need for complementary policies that would improve the accuracy of employers’ perceptions. No one would argue that the program is the only policy instrument that should be used to increase employment for people with criminal records. It should instead be seen as one piece of a puzzle.

While the above points represent limitations of insurance policies, they should not be seen as weaknesses. It is not a weakness of the program that it fails to alter employers’ perceptions, because that is not its purpose. Increasing employment in groups with labour-market challenges is a complex problem with many issues at play. Policies aimed at improving the accuracy of employers’ perceptions are bold attempts to change hearts and minds. They are likely to take time to have an effect and may never fully do so. In the meantime, other policy levers are required to incentivize employers’ actions, encouraging them to hire from these populations at a more optimal rate. That is what the bonding program seeks to do.

At the same time, we should not be too quick to dismiss the capacity of employer incentives to have an indirect impact on the perceptions of risk associated with hiring people with criminal records. If a bonding program, especially an expanded version of the current US program, is successful in encouraging more employers to hire more of this population, then we would expect more workplace interactions between employers and employees with criminal records. More familiarity with this population and more experience working with them should, over time, reduce employers’ concerns over hiring them.

Canada does not currently offer anything similar to the US program. The Government of Ontario offers a small wage subsidy through the Skills Development Fund, and the Government of British Columbia offers wage subsidies for businesses that are clients of the WorkBC Agency. 34 34 Ontario Ministry of Labour, Immigration, Training and Skills Development, Skills Development Fund Training Stream Application Guide (July 2024), https://forms.mgcs.gov.on.ca/en/dataset/on00723; British Columbia Ministry of Social Development and Poverty Reduction, WorkBC Wage Subsidy: Employer Application Guide (June 2023), https://www.workbc.ca/media/1528/download?inline=. These are relatively small programs, however.

As discussed in the previous section, a bonding program does not constitute a complete strategy for employing more people with criminal records. The relative ease with which such a program could be implemented, though, makes it a good first step toward encouraging greater employment of people with criminal records, as well as educating employers, citizens, and civil-society institutions about the need for a more comprehensive strategy.

Because a bonding program could be introduced by any level of government and the Canadian federation is diverse, a one-size-fits-all strategy is not advisable. The framework below outlines the key considerations that each government would need to engage.

One level of government could implement the program, or several levels could work together. Although a bonding program would better fit the federal and provincial governments, a municipal government could conceivably operate one as well.

The US program was initiated and is funded by the federal government, but it is operated in each state by a coordinator who is an employee of the state government.

The US program provides bonds between $5,000 and $25,000 (although these amounts have not increased in recent years). As of November 2024, the equivalent in Canadian dollars would be about $7,000 to $34,000.

As in the US, most bonds would be issued for the minimum amount. Only jobs with greater responsibility, such as for company resources, would require the larger amounts. A government could include the criterion that the value of equipment or financial resources being handled in the job should determine the amount of the bond. As discussed below, though, this criterion should be kept simple so as not to complicate the application process.

Governments might wish to consider bonds above the $7,000 to $34,000 range. Given that only 1 percent of bonds end in a claim in the US program, the cost of issuing larger bonds would probably be negligible. That said, the Canadian-dollar equivalent of the US program’s range would be a good place to start.

At minimum, the Canadian program should have the same scope as the US program, namely, covering losses from theft, forgery, larceny, fraud, and embezzlement committed by the employee. It would also be prudent to adopt the US program’s length of coverage, establishing a six-month term, renewable once for each employer.

People with criminal records have widely researched challenges entering the labour market, and people in recovery from substance abuse have been the subject of increasing attention from Canadian governments, given the current drug crisis. These two groups would be a good starting point, although other groups could be included as well.

The cost would be determined by a tendering process and further negotiation between the government and an insurance provider. The government could provide the insurance itself, but it makes more sense to contract it out, as the product fits with the nature of an insurance company. In either case, the government must bear the cost of the insurance, as the whole purpose of the program is to shift the cost of the risk away from the employer.

Simplicity of administration is key. For the program to incentivize employers, it must not add any significant complexity or delays to their hiring processes. The program should be as easy as possible for both employees and employers to participate in.

The bond applications should be approved routinely. The fact that the bonds are so cheap and low-risk means that a bureaucracy with more than the minimum amount of scrutiny would simply not be worth the cost. Approval could occur after a simple attestation that the applicant has a criminal record, with minimal auditing of a small number of cases.

As such, the program should have very low overhead. As in the United States, the program may even be able to operate with just one or two staff. As the program grows, additional staff could be added as necessary.

The US experience suggests that the program will be underused unless widely advertised. Governments should therefore dedicate a generous percentage of the program’s budget to advertising, especially in the early years, to ensure awareness among employers, employment agencies, non-profit organizations involved in the reintegration of people with criminal records, and other civil-society organizations involved in this area. Governments could also distribute materials to job applicants that they can provide to prospective employers during the application process. This would ensure that the employers are aware of the program at the very moment of considering the applicant.

The US program collects data on the person being bonded (sex and ethnicity), the job (occupation, industry, wage, and hours of work per week), and the employer (size of the company and whether it is for-profit, not-for-profit, or government). 35 35 U.S. Department of Labor, Advisory: Training and Employment Guidance Letter No. 15-18. Governments in Canada may be interested in collecting this and additional information, provided it does not significantly add to the paperwork burden.

The cost to the government’s treasury would be minimal. A program like this is also scalable to any maximum level of funds that the government would be willing to provide, as the US federal government did in providing $5 million to the states in 2019.

As a starting point, a Canadian government could provide an amount equivalent to the US government’s expanded program, according to the population of its jurisdiction. Thus, if the federal government created the program, a reasonable starting budget would be about $800,000. If the Ontario government created it, it could start with a budget of $300,000. Quebec could start with $200,000, Alberta with $100,000, and so on. 36 36 These figures are calculated by multiplying the national amount of $800,000 by the population ratio according to Statistics Canada’s population estimates for the first quarter of 2024, and rounding to the nearest $100,000. Table: 17-10-000901: Population Estimates, Quarterly (December 17, 2024), https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=1710000901.

If we assume that Canadian bonds would cost about as much as US ones, a $7,000 bond would cost approximately $140. Thus, with $800,000, a made-in-Canada program could provide a bond to nearly 6,000 people across the country. Given that nearly four million Canadians have a criminal record, there should be more than enough demand for that number of bonds. 37 37 Public Safety Canada, Criminal Records (last modified March 18, 2020), https://publicsafety.gc.ca/cnt/trnsprnc/brfng-mtrls/trnstn-bndrs/20191120/017/index-en.aspx. The actual population that would benefit from this program is even larger, as this figure does not include members of other groups that the program might also target. As the program becomes more well-known and demand for it increases, governments could increase the funds available accordingly.

Costs relating to personnel and advertising might push a national program up to $1 million, if the program employed one full-time staff person at the average federal public service cost of about $125,000 and spent $75,000 on advertising. 38 38 Office of the Parliamentary Budget Officer, Canada, Personnel Expenditure Analysis: Update (April 4, 2023), https://www.pbo-dpb.ca/en/publications/RP-2324-002-S--personnel-expenditure-analysis-update--examen-depenses-personnel-mise-jour. While the ratio between the cost of administration and the cost of the bonds themselves might seem high, it is important to remember that this is because the bonds themselves cost so little, at $140 per $7,000 bond. The proper measure of effectiveness should be the ratio of the number of staff to the number of people who are bonded. By this measure, the cost of administration is very low.

The cost of this program must be compared to alternative ways of incentivizing employers to hire people from the targeted groups. As stated above, a government would need to introduce an 80 percent wage subsidy to have the same effect as a bonding policy like the one contemplated here. To cover 80 percent of the wages of 6,000 full-time workers earning minimum wage for six months, a government would need to budget almost $80 million. 39 39 Assuming full-time work at Ontario’s minimum wage as of May 2024 ($16.55 per hour), the earnings of 6,000 people over six months would equal approximately $99 million. A wage subsidy covering 80 percent of that would cost the government approximately $79 million. Even a much more modest wage subsidy like the Work Opportunity Tax Credit in the United States—which covers up to 40 percent of wages up to a maximum subsidy of $2,400 per worker—could cost almost $19 million to cover 6,000 workers. 40 40 Internal Revenue Service, Work Opportunity Tax Credit (August 31, 2023), https://irs.gov/businesses/small-businesses-self-employed/work-opportunity-tax-credit. The maximum $2,400 translates to $3,288 in Canadian dollars. Assuming that employers receive the maximum subsidy for every eligible worker, the program would cost $19.7 million to cover 6,000 workers.

Finally, it must be noted that the cost of the program would be repaid quickly through income taxes paid by the program’s beneficiaries, who might not otherwise have received employment income. There are few programs that can truly be said to pay for themselves. This is one of them.

Governments in Canada should do more to encourage the hiring of those who have difficulties getting jobs, including those with criminal records. The US Federal Bonding Program offers one model for doing so.

A made-in-Canada bonding program would be a simple, low-cost incentive that would directly target a core reason why employers say they are reluctant to hire from this population, that is, the risk of suffering losses from theft. The positive externalities associated with jobs for people with criminal records, and the fact that the actual risk is lower than employers perceive it to be, suggest a market failure that can be corrected by government taking the risk upon itself. And the risk is relatively low, which makes a bonding program such as the one contemplated in this paper a very cost-effective way of increasing the employment of this group.

A bonding program is not the ultimate answer to the problem of low employment levels for people with criminal records or in recovery from addictions, but it could go a long way. No program is a silver bullet. This, like any policy, should be seen as one tool in the policy toolbox.

At the same time, we should not underestimate how the discussion that would accompany the implementation of a bonding program would itself benefit the employment prospects of people with criminal records. Policies to support this population’s re-entry into Canadian society can be controversial, but this policy would be very low-risk for policy-makers who recognize the need for action in this area. As such, this policy could be a useful on-ramp for making the case to citizens that pro-employment policies are an important part of a criminal-justice agenda.

Even more fundamentally, however, we should not underestimate the impact that this program could have in the individual lives of people who face challenges entering the labour force. A bond could be just the assurance an employer needs. With that assurance, the employee has the opportunity to earn an income, develop and share their skills with their communities, and leave their past mistakes behind. The bond is one way of opening a door to that opportunity.