John Deutsch Institute

Zach Robb

JDI Policy Insight: On Fairness in AI-Assisted Decisions

August 5, 2026

Kingston, Canada – August 2026 – The John Deutsch Institute (JDI) has released a new JDI Policy Insight paper, Fairness in AI-Assisted Decisions: Five Insights for Policymakers, written by Tanvir Ahmed Khan, a PhD candidate in the Department of Economics at Queen’s University. Algorithms now inform decisions about credit, hiring, benefits, healthcare, and pretrial release. The paper argues that the question policymakers usually ask about them is the wrong one.

That question is whether the algorithm is fair. In most high-stakes settings the algorithm does not make the decision: it produces a score that a loan officer, case worker, recruiter, or judge then interprets. The outcome depends on the model and on how people and institutions use it. Canadian federal guidance already reflects this, since the Treasury Board’s Directive on Automated Decision-Making covers tools that support human decision-makers rather than replace them.

Khan draws five points from the literature. Competing definitions of fairness can conflict, and the conflict is sometimes mathematical rather than political: the dispute between ProPublica and Northpointe over the COMPAS recidivism tool arose because a score can be calibrated across racial groups while still producing different error rates, when underlying base rates differ. A more accurate model does not guarantee smaller disparities, since sharper predictions filtered through biased human cutoffs can widen gaps rather than close them. A disparity that grows after adoption does not reveal its own cause, which might be the training data, human overrides, or a shifted threshold. How fairness risks are framed changes what organizations do about them, with experimental evidence showing that managers told algorithmic bias is inescapable become more likely to abandon the tool for unguided human discretion. And evaluating a model is not the same as evaluating the decision process built around it.

Khan recommends:

  1. Specify the fairness objective before deployment. Decide in advance whether the priority is equal treatment of similar cases, parity in error rates, or equitable access, since post-deployment review is hard to interpret without a stated benchmark.
  2. Evaluate the decision process, not just the model. Assessment should trace the full chain from training data through thresholds, override rules, and exception handling to the final allocation.
  3. Track how human discretion changes after adoption. Log when recommendations are overridden, by whom, and why, and review how override patterns are distributed across groups.
  4. Treat model upgrades as governance events. Better prediction can expose institutional bias that noise previously obscured, so a material performance change should trigger renewed review of the surrounding procedures.
  5. Communicate fairness risks without fatalism. Internal reporting should separate identifying a risk from concluding the system should be abandoned, and state the benchmark being used.

Khan’s assessment is that Canada’s existing architecture already recognizes the decision process as the policy object, but does not yet ensure departments generate evidence about the operational margin where scores become decisions. Closing that gap is the next stage of responsible AI governance.

The full paper is available as JDI Policy Insight 26-0802 on the JDI website at jdi.queensu.ca.

About the John Deutsch Institute and the JDI Policy Insight series
The JDI at Queen’s University conducts rigorous, policy-relevant economic research to inform decision-makers in government, industry, and civil society. Through events, publications, and collaborations, the Institute fosters evidence-based dialogue on critical economic challenges facing Canada and the world. The newly launched JDI Policy Insight series provides balanced academic summaries of key policy issues, helping readers assess the benefits, costs, and risks of reforms and understand what we know and don’t know about policy.

JDI Policy Insight: On Employment and Health

August 5, 2026

Kingston, Canada – August 2026 – The John Deutsch Institute (JDI) has released a new JDI Policy Insight paper, Employment and Health: Mechanisms Beyond the Empirical Studies, written by Jin Young Yoon, a PhD candidate in the Department of Economics at Queen’s University. Canada spends roughly 11 to 12 percent of GDP on health care and substantial additional sums on Employment Insurance and active labour market programs, yet the two policy areas are usually designed separately. The paper asks what connects them.

The obstacle to answering that question is that employment and health each influence the other. Healthier people are more likely to find and keep work, which makes it hard to read a simple correlation as evidence that work improves health. Yoon approaches the problem by examining the specific channels through which employment might operate, using two studies with very different findings as a way in.

The Whitehall Study, which followed more than 10,000 British civil servants from 1967, found that health improved steadily with employment grade: lower-ranked staff showed more angina, ischemia, and chronic bronchitis than their senior colleagues, despite all of them being employed. The Community Employment Innovation Project, a Nova Scotia experiment that randomly assigned three-year community jobs to half of 1,514 unemployed participants, found that employment improved health, with effects persisting for 14 months after the jobs ended.

Yoon argues the two results are consistent once the mechanisms are separated. Income matters, and both studies involve gains in it. But so do working hours, with UK civil servants averaging 45 hours a week against 35 for CEIP participants. So does perceived job control, which is low in the junior ranks of a hierarchical bureaucracy and higher in the varied community roles CEIP created. So does workplace social interaction, which the community-based design of CEIP encouraged and routine civil service work did not. Employment can improve health or damage it, depending on which of these channels dominates.

Yoon recommends:

  1. Evaluate labour programs on health outcomes. Job creation and re-employment initiatives should be assessed on participant health alongside employment rates and earnings, since the jobs they provide can improve or worsen it.
  2. Regulate for job quality, not just job quantity. The Whitehall results show that employment alone does not protect health, pointing to a role for standards governing autonomy, working hours, and fair compensation.
  3. Preserve the non-monetary benefits of work. Because unemployment removes social interaction as well as income, unemployment policy should attempt to sustain social engagement and continuing education.
  4. Tailor policy to the target population. For people moving out of unemployment the gains come through income, routine, and social contact; for those already in low-quality jobs, working conditions matter more than employment status.

Yoon is careful about what the evidence cannot yet settle. The CEIP effects are measured over months rather than years, and whether health gains persist under sustained employment remains open, as does the question of which job characteristics matter most.

The full paper is available as JDI Policy Insight 26-0801 on the JDI website at jdi.queensu.ca.

About the John Deutsch Institute and the JDI Policy Insight series
The JDI at Queen’s University conducts rigorous, policy-relevant economic research to inform decision-makers in government, industry, and civil society. Through events, publications, and collaborations, the Institute fosters evidence-based dialogue on critical economic challenges facing Canada and the world. The newly launched JDI Policy Insight series provides balanced academic summaries of key policy issues, helping readers assess the benefits, costs, and risks of reforms and understand what we know and don’t know about policy.

Five New JDI Policy Insight Papers Released

August 5, 2026

Kingston, Canada – August 2026 – The John Deutsch Institute (JDI) has released five new JDI Policy Insight papers this month, each written by a PhD candidate in the Department of Economics at Queen’s University. The papers span immigration, consumer credit, platform regulation, AI governance, and the relationship between work and health.

This month’s papers include:

Together, the three papers reflect the range of the JDI Insight Series: close, evidence-based looks at policy questions that affect Canadians directly, written by the next generation of Queen’s economists. All three are available on the JDI website at jdi.queensu.ca.

JDI Policy Insight: How Immigration Impacts the Economy

August 1, 2026

Kingston, Canada – July 2026 – The John Deutsch Institute (JDI) has released a new JDI Policy Insight paper, Understanding How Immigration Impacts the Economy, written by Luke Rawling, a PhD candidate in the Department of Economics at Queen’s University. The paper reviews decades of research on immigration’s economic effects and asks what happens to existing workers when a country admits more immigrants.

Answering that question is harder than it appears. Immigrants tend to settle in cities where the economy is already doing well, which makes it difficult to separate immigration’s effect on wages from the local conditions that attracted immigrants in the first place. Rawling reviews the methods economists have developed to get around this problem and makes the case that the most reliable estimates come from studies built around a single, unanticipated shock.

Those studies do not point in the same direction. After the 1994 peso crisis sent hundreds of thousands of Mexican migrants into the United States, wages for low-skill workers fell by as much as 1.4 percent in the first year, though the effect had mostly faded within five. A similar inflow of Czech commuters into Germany barely moved wages but cut native hiring sharply. In Denmark, refugee arrivals had no negative effect on low-skilled natives at all, and appear to have moved them into better-paying, less manual work.

Rawling treats this variation as the central result. Immigration adds to the labour supply, but it also spurs innovation, supports the creation of new firms, and produces complementarities that raise demand for existing workers. Which of these forces dominates depends on local economic conditions, the skills immigrants bring, and how individual firms choose to respond.

For Canada, where immigration levels are set by federal and provincial governments rather than by employers, that has practical consequences. Immigration has been at record highs while key sectors continue to report labour shortages, and immigrants remain underrepresented in construction and public administration, two of the sectors under the most strain. Rawling recommends:

  1. Align targets with private-sector demand. Immigration targets should reflect the labour needs firms actually report, since firms are where new arrivals are ultimately absorbed.
  2. Prioritize business-creating entrepreneurs. Selection should give more weight to applicants who can show investor backing, since immigrant-founded firms create jobs as well as fill them.
  3. Rebalance sectoral selection. Tools such as category-based selection and the Federal Skilled Trades stream, which currently accounts for about 0.5 percent of economic immigrants, could bring more newcomers into sectors where added labour would relieve pressure on housing.

With Canada’s labour force growth increasingly dependent on immigration, Rawling argues that these design choices matter as much as the headline admission numbers.

The full paper is available as JDI Policy Paper 26-0703 on the JDI website at jdi.queensu.ca.

About the John Deutsch Institute and the JDI Policy Insight series
The JDI at Queen’s University conducts rigorous, policy-relevant economic research to inform decision-makers in government, industry, and civil society. Through events, publications, and collaborations, the Institute fosters evidence-based dialogue on critical economic challenges facing Canada and the world. The newly launched JDI Policy Insight series provides balanced academic summaries of key policy issues, helping readers assess the benefits, costs, and risks of reforms and understand what we know and don’t know about policy.

JDI Policy Insight: On the Credit Card Market

August 1, 2026

Kingston, Canada – July 2026 – The John Deutsch Institute (JDI) has released a new JDI Policy Insight paper, Perceptions Versus Realities in the Credit Card Market: Consumer Vulnerability, Competition, and Inequality, written by Chi Danh Dao, a PhD candidate in the Department of Economics at Queen’s University. With Canadian credit card debt above CAD 122 billion, the highest level since 2007, the paper examines how well the evidence supports common claims about that debt.

Dao begins with a measurement problem. Credit bureau data indicate that 48 percent of Canadian cardholders carry a revolving balance, while Bank of Canada surveys put the figure at 26 percent. The two sources capture different populations: surveys tend to reach persistent revolvers who carry debt month after month, while credit reports also include cardholders who usually pay in full but occasionally face an unexpected expense. Quebec’s 2019 reform shows why the distinction matters. Raising the minimum payment from 2 to 5 percent lowered total revolving debt but increased delinquency, since the rule that helped one group tightened constraints on the other.

The profile of revolving cardholders also complicates the standard account. They are often described as financially distressed, yet they report asset holdings and levels of financial comfort comparable to cardholders who pay in full, roughly 3 percent are 30 or more days delinquent, and they are 33 percent more likely to hold a mortgage.

On competition, Dao argues the usual intuition does not hold. Networks compete for consumers, who can switch easily, by offering richer rewards, and recover the cost from merchants, who cannot afford to turn cards away. Canadian interchange fees reach 2.5 percent of a transaction against 10 to 35 cents for a debit tap, and 89 percent of businesses accept credit cards regardless. Federal remedies have had limited effect for the same reason: negotiated fee reductions left rewards intact, and only about 12 percent of retailers planned to use their new surcharging rights.

On inequality, convenience users capture an average of CAD 192 in net rewards annually while revolvers face an average net cost of CAD 672. Dao argues net rewards are an incomplete measure, since they count interest paid without counting the value of the credit it buys. About 2 percent of Canadian consumer spending is financed through revolving debt. Measured through consumer surplus, credit cards still worsen inequality, though by considerably less than the headline figures suggest.

Dao recommends:

  1. Match the data source to the policy question. Rules such as minimum-payment requirements affect persistent and occasional revolvers in opposite ways, and the dataset a regulator relies on determines which group is visible.
  2. Pair fee reforms with demand-side measures. Since competition can raise merchant fees, fee caps and surcharging rights are unlikely to reduce overutilization without addressing consumer demand.
  3. Measure inequality through consumer surplus. Assessments of distributional effects should account for the value of unsecured credit access alongside rewards, interest, and fees.

The last point bears on a live proposal. Following legislative moves in the United States, there is renewed debate about capping Canadian credit card interest rates. A cap would lower borrowing costs, but Dao cautions that banks may respond by reducing credit limits. Higher-income households have other borrowing options; lower-income cardholders may have to cut consumption instead, worsening the inequality the cap was meant to address.

The full paper is available as JDI Policy Insight 26-0702 on the JDI website at jdi.queensu.ca.

About the John Deutsch Institute and the JDI Policy Insight series
The JDI at Queen’s University conducts rigorous, policy-relevant economic research to inform decision-makers in government, industry, and civil society. Through events, publications, and collaborations, the Institute fosters evidence-based dialogue on critical economic challenges facing Canada and the world. The newly launched JDI Policy Insight series provides balanced academic summaries of key policy issues, helping readers assess the benefits, costs, and risks of reforms and understand what we know and don’t know about policy.

JDI Policy Insight: Making Platforms Pay for News

August 1, 2026

Kingston, Canada – July 2026 – The John Deutsch Institute (JDI) has released a new JDI Policy Insight paper, Making Platforms Pay for News: Policy Design and Outcomes, written by Seth Kushniryk, a PhD candidate in the Department of Economics at Queen’s University. The paper examines three attempts to address the same problem, that platforms profit from news content while publishers carry the cost of producing it, and considers what the outcomes suggest about how such policies should be designed.

The underlying trend is well documented. Online advertising spending in Canada is projected to reach $21.9 billion in 2025, an 80 percent increase over five years, with Google and Meta accounting for most of the growth. Over the same period Canada lost 11 percent of its print outlets, and roughly 2.5 million Canadians now live in areas served by one local news outlet or none. The three jurisdictions responded in different ways:

  • Australia made bargaining mandatory but relied on the threat of enforcement rather than enforcement itself. The Treasurer retained discretion to formally designate a platform and never used it. The possibility of binding arbitration was still enough to produce roughly 30 agreements worth an estimated AUD$200 million.
  • The European Union built its framework on copyright, granting publishers a right covering online uses of their content. Platforms could sidestep it by not displaying excerpts. French publishers eventually secured payment through the competition regulator, which ordered Google to negotiate and fined it €500 million.
  • Canada combined the two approaches, pairing mandatory bargaining with an exemption pathway. Google took the exemption and agreed to contribute CAD$100 million annually through a collective representing more than 1,400 news businesses. Meta blocked news for Canadian users in August 2023 rather than negotiate.

The consequences of Meta’s withdrawal are the clearest evidence in the paper. Canadian outlets lost 85 percent of their engagement on Facebook and Instagram, and within a year 215 news organisations had stopped posting on social media altogether, 98 percent of them local. Only 22 percent of Canadians know the ban exists, so most users never sought out substitutes.

Kushniryk concludes that the credibility of the enforcement threat matters more than whether a policy is grounded in copyright or competition law, and that platforms value news differently enough that a single set of rules will not work on all of them. Kushniryk recommends:

  1. Design for platform heterogeneity. Search engines depend on news and can bear mandatory arbitration, while social media platforms, where news is more easily substituted, may warrant lighter requirements.
  2. Build in distributional safeguards. Specify minimum allocations for local and small publishers. In Australia, major outlets secured deals worth tens of millions while 160 regional newspapers together received under AUD$1 million.
  3. Require public reporting of deal terms. Australia’s confidentiality provisions make it impossible to assess whether funds reached public interest journalism. Canada’s recipient-level disclosure offers a better model.
  4. Plan for withdrawal before enacting. Identify which outlets depend most on platform distribution and prepare funding and communication strategies in advance.

Kushniryk’s broader conclusion is that bargaining codes work best alongside other measures. They can secure meaningful payments from platforms that rely on news, but they have limited reach over platforms that do not, and they offer no guarantee that funding will flow to the local journalism most at risk.

The full paper is available as JDI Policy Insight 26-0701 on the JDI website at jdi.queensu.ca.

About the John Deutsch Institute and the JDI Policy Insight series
The JDI at Queen’s University conducts rigorous, policy-relevant economic research to inform decision-makers in government, industry, and civil society. Through events, publications, and collaborations, the Institute fosters evidence-based dialogue on critical economic challenges facing Canada and the world. The newly launched JDI Policy Insight series provides balanced academic summaries of key policy issues, helping readers assess the benefits, costs, and risks of reforms and understand what we know and don’t know about policy.

  • Recent Posts

  • Recent Comments

      • Privacy/Credits
      Queen’s University is situated on traditional Anishinaabe and Haudenosaunee Territory.
      @ 2020 Queen’s Law. All Rights Reserved.

      slot server thailand

      slotter88