Research
Selected Working Papers
(1) The Labour Supply Effects of Public Pension Reform in Canada (Job Market Paper)
You can access the PDF file here
Abstract:
Focusing on reforms to Canada’s public pensions in 2012/13, I study retirement and public pension take-up decisions, and associated interactions among elements of the public pension system. Given increasing longevity, the reforms’ main objective was to incentivize labour supply among the population over age 60. Unlike most other developed countries, Canada did not increase its statutory public pension eligibility ages. Instead, for the contributory Canada Pension Plan (CPP) these reforms broke the link between take-up and retirement decisions, allowing individuals to receive benefits without retiring while simultaneously increasing the incentive to defer public pension take-up regardless of retirement status. At early eligibility ages, between 60 and 64, I estimate that the reforms delayed retirement and reduced the retirement spike at age 60. While I do not observe a decline in the retirement hazard spike at age 65, I find an increase in the duration of public pension deferral beyond age 65 for those who had not yet taken up benefits. The CPP reform interacted with the non-contributory Guaranteed Income Supplement (GIS) portion of the public pension system for low-income families by placing an exemption threshold for paying CPP premiums at the same earnings level as the GIS earnings exemption. For many, this increased the effective tax rate above the exemption. The earnings bunching at the exemption grew markedly, however it is only evident among select industries and provinces suggesting policy responses among employers. The reforms also introduced a deferral option for the income tested non-contributory Old Age Security (OAS) program. I observe downward kinks in earnings and employment beyond the OAS threshold preceding the deferral change. Once tax filers had the option to defer, however, the magnitude of the kinks decreased and a kink in OAS take-up developed. Disentangling the retirement and public pension take-up decisions is an effective approach to supporting delayed retirement without some of the downsides associated with increasing statutory ages, but evaluations of such reforms need to consider interactions among various aspects of the public pension system.
(2) Low-Income Support Programs and the Retirement Decision of Couples in Canada
Abstract:
I study the effects of Canada’s major federal low-income support program for low-income couples on the labour supply decisions of those aged 60 and older. Considering a 2008 reform that increased the Guaranteed Income Supplement (GIS) earnings exemption, an earnings threshold above which labour income induces a tax-back of the GIS benefit, I estimate the impacts of reducing the participation effective tax rate on the labour supply decisions of low-income couples. I compare labour supply outcomes of those aged 65-70 in an event-study where treatment status is defined by earnings-history. For those in the lowest earnings-history decile, my results suggest that the reform delayed retirement but had no impact on earnings. The reform also applied to the Spouse’s Allowance (SPA), which is a component of the GIS program available to the spouse of a GIS recipient aged 60-64. In an analysis of couples’ joint retirement decision making, I employ an event-study that compares labour supply outcomes of GIS-eligible couples between those who are age-eligible for the SPA benefit to those who are not. While my results suggest that the reform delayed retirement and increased the earnings of SPA recipients, they indicate negligible labour supply effects for the older GIS-eligible spouse aged 65-69. However, a 2011 reform to “top-up” the GIS benefit for exceptionally low-income couples with little income beyond GIS receipt increased both the benefit payout and the GIS tax-back rate on earnings by 25 percent. The top-up is observed to reduce the labour supply for individuals in the bottom earnings-history decile but had no effect on the labour supply of couples in which one member is SPA-eligible. Overall, my results indicate that while financial incentives through low-income support programs for retirees affect individuals’ labour supply decisions, their impacts vary depending on couples’ circumstances.
(3) Residency-Based Social Pensions and Retirement Timing in Canada
Abstract:
I use administrative tax filer data and a discrete-time mixed proportional hazard model to study the retirement and public pension take-up timing of late-age-arrival immigrants to Canada compared to non-immigrants. I find that late-age-arrival immigrants, who must satisfy ten years of permanent residency to be eligible for a prorated Old Age Security (OAS) pension, are more likely to be employed at ages 60-70 than non-immigrants but are also more likely to take-up Guaranteed Income Supplement (GIS) benefits, an indication of exceptionally low family income in retirement. However, I find that after satisfying the residency requirements, the difference in the late-age-arrival immigrant to non-immigrant employment rate at ages 65 and over becomes statistically indistinguishable from zero.