Guaranteed Annual Income

On Wed, January 15th at 8pm EST there will be a Twitter discussion about the idea of a basic income supplement with economists Stephen Gordon (Laval), Kevin Milligan (UBC), Mike Moffatt (Ivey) and myself facilitated by the twitter account, @bis_pilot. To kick off that discussion, this blog post will provide some background information on this policy idea.

A basic income supplement, better known as a Guaranteed Annual Income (GAI), provides a floor level of income, usually set to a level to alleviate poverty. There are two forms of a GAI: Negative Income Tax (NIT) and Universal Demogrant (UD). The NIT involves the payment of money to those earning below a specified level, no taxes paid by those above that but below another threshold, and finally those earning above the second threshold pay tax on that income above the threshold. The UD is instead a non-taxable transfer to all citizens. The NIT is the more popular version of the GAI, especially among economists, probably because it is more targeted to the poor and far less costly than a UD.

A GAI is not a new idea by an means. As far as I can tell, it dates back to 18th century England and the Speenhamland system, which was intended to mitigate rural poverty, and later became the Elizabethan Poor Law. This law set a basic income floor that was scaled to the price of bread needed to sustain a given family size

In modern times, Milton Friedman brought the notion of a GAI back to the policy agenda with his argument for a negative income tax (NIT). His work spurned a series of NIT experiments in the US between 1968 and 1979. The first was the New Jersey Income Maintenance Experiment that included just under 1400 families. This was followed by the Rural Income Maintenance Experiments in Iowa and North Carolina. Two others took place in Gary, Indiana as well as Seattle and Denver. This last one was in operation until 1982. The main findings of the experiments was that there did exist a work disincentive.

The idea of a GAI is not at all new in Canada either. In 1974, we ran the Mincome pilot in Manitoba. This was a collaboration between the federal government and the province on a pilot program to assess the affect of a GAI on work incentives. It was a 3-year pilot that included 1300 families and had 3 different GAI levels. The key finding was that work disincentives were modest, especially compared with the data from the US: there was some reduction in work hours, but predominantly among female breadwinners with young children. This demonstrates the need to not rely on American evidence of policy matter and collect our own evidence regarding policy implications (similar to that regarding minimum wage).

Not only did Canada run that avant garde experiment but it comes as a surprise to many Canadians that we actually have many GAI like program. Many of our current programs are a form of GAI, including the OAS, GIS, CPP, CCTB, EI, and social assistance.

The GAI has been advocated by many in recent years, including Senator Hugh Segal. Many argue it is a preferred policy to minimum wage or in-kind transfers as it gives money to those most in need and does not rely on private businesses to deliver on social policy. Canada has had great success with using GAI policies to dramatically reduce senior poverty. Such policies could be used to do that same for child poverty, which has not seen any decline over the same period.

The are two main criticisms of the GAI. The first involves the work disincentives, but the evidence is mixed on this feature and depends on the implementation of the GAI. The second is the cost. GAI can be costly, especially if the UD form of GAI is chosen. If instead, existing programs are consolidated and it is implemented as a form of NIT, then the cost is much lower. The cost could also come from eliminating many of our boutique tax credits that disproportionally benefit the rich, as I have written about before.

If this is a topic that is of interest to you, join our discussion on Wednesday and see what it is all about.