(with Kourtney Koebel)
Many scholars have been highly critical of the newly elected Ontario government’s announcement to cancel the basic income pilot currently underway in the province, claiming it is a wasted opportunity to collect data on an important policy issue.
While preserving the data that has been collected to date for future research purposes is important, we are of the opinion that the pilot was unlikely to tell us anything we don’t already know from theory and previous empirical research about the probable effects of a basic income on recipients’ outcomes. There are previous studies in both Canada and other countries on the positive outcomes of various basic income pilots, and also on permanent cash-transfer policies and programs that tell us something about what we might expect.
Like many academics who span a wide range of disciplines and ideological spectrums, we are supporters of a basic income guarantee for evidence-based efficiency and equity reasons that are too numerous to outline here. However, we’re not supportive of basic income pilot experiments. Here’s why:
First, the reality of democracies is that changes in government are almost always associated with some reversals of the previous government’s policies. Pilots are even more susceptible to this than actual legislation or social programs, and the previous government implemented the pilot project too late to guarantee its completion. Other basic income pilots that have taken place in Canada – one in Dauphin, Manitoba in the 1970s and, more recently, the 2017 pilot in Ontario, ended due to changes in government. Of course, any social policy is vulnerable during a government transition. For example, child benefits in Canada have changed substantially since the first Family Allowance program came into existence in 1945. However, the relative ease of eliminating something experimental puts pilots – and the individuals participating in them – at a higher risk.
Second, while we applaud the increasing influence of behavioral research and evidence-based approaches on improving public policy interventions, both academics and policy-makers have become overly enamored with the idea of randomized controlled trials (RCTs) as the “gold standard” to evaluate public policies, without fully acknowledging the difficulties associated with ensuring internal and external validity in this particular context.
There were already methodological issues with the pilot that would have made the results suspect due to a lengthy application package that likely led to serious selection bias in who was motivated to apply for the program. Moreover, the pilot’s limited duration raises critical questions about whether participants would change their behavior in the same way as they would if the program was viewed as relatively permanent. The data collection associated with the pilot and interference by politicians and advocacy groups may also affect how recipients respond to the pilot – potentially in ways they otherwise wouldn’t if they weren’t being monitored.
The exclusion of the labour supply responses of those who would be “net contributors” to funding a basic income in the pilot further undermines its external validity as a way to test the feasibility and sustainability of a comprehensive program. In addition, because the basic income pilot is limited to low-income individuals and workers, researchers would not be able to make any conclusions about a “universally accessible” basic income, as middle and high-income individuals were ineligible to participate. Even if they would not have received the basic income, the opportunity to obtain it should have been available. Excluding individuals who experience a change in income limits its applicable relevance to a true basic income. Moreover, the small size of the sample relative to the broader labour market also compromises the external validity of the pilot results.
We are not alone in our reticence or are even the first to voice these concerns about basic income pilots. One of the leading worldwide proponents of a universal basic income, Philippe Van Parijs, states in his 2017 tome with Yannick Vanderborght on this topic that “experiments are not very promising in terms of what we can learn from them about the real-life sustainability of basic-income schemes…we must try to infer causal links from correlations...[using] econometric models that claim to predict…what would happen if a basic income were introduced (p. 144).” That is, a permanent basic income is likely to have effects that cannot feasibly be explored in a pilot.
Finally, and perhaps most important, Ontario’s basic income pilot was ultimately unsustainable as a provincial program, and this is important to acknowledge in light of other provincial announcements to pursue basic income studies and pilots. Any affordable implementation of an adequate basic income requires tax reform and both federal and provincial contributions, and any sustainable program requires inter-jurisdictional collaboration on development and agreement of a framework between the federal government, all provinces and First Nations communities. Provinces, on their own, could finance a very modest basic income, but to realize the full guarantee that many supporters advocate, federal-provincial coordination is necessary, as well as the opportunity for participation by interested First Nations communities in the development of a framework.
The necessity for collaboration between both levels of government creates additional issues for running a provincial pilot, as seen by PEI's unsuccessful attempt to establish a partnership with the federal government to run a basic income pilot. Federalism is still likely the biggest political barrier to its implementation and sustainability, and a pilot will tell us very little about either. The only way to identify this is by trying to actually implement a basic income, which requires nothing more than political will.
In this regard, there is also serious political risk associated with pilots. While pilots can raise public awareness, which is good, a failed or poorly designed pilot could be the nail in the coffin for future implementation of a full basic income guarantee. Moreover, pilots can be used by politicians to put off calls for the immediate implementation of necessary changes to current social policy and programs, which also negatively impacts all the people who aren’t selected to participate in the pilot.
Even if the current Ontario government is ideologically opposed to a universal basic income, Ontario has elected this government for the next four years, and there are second best alternatives. Scholars and advocates would do well to focus their attention and energies on providing the government with some of these options. During the election campaign, Ford discussed a variation of a negative income tax scheme, by proposing the elimination of taxes for minimum wage workers, and providing them with a refundable tax credit. While this does not address our major concerns about the inadequacy of supports for working-age Canadians and the problems associated with current social assistance programs, it does show that the government may be willing to consider proposals that help low income people and take seriously their concern about work disincentives. While we think the government’s concerns about work disincentives and problems with labour market attachment under a basic income guarantee are overstated - the view is solely based on the neoclassical conceptualization of work and why people work - we do think the concerns are still legitimate and need to be taken into account in designing the most effective program possible.
In our view, the lack of clarity around how current participants will be treated in wake of the pilot cancellation is the most pressing issue for the government to address in the short term, especially because the pilot participants are low income individuals and households. Canceling the formal pilot while continuing payments to all participants until the original proposed end of the pilot, and then creating a transition plan for these individuals and households would be the most ethical way to address this concern.