An old truism among economists is: Whatever you subsidise, you get more of.  An illustrative instance can be seen by comparing Maine and New Brunswick, adjacent jurisdictions with similar economies: Both are based on fishing, forestry, and tourism; both have incomes below the respective national average; both experience low population growth.

Yet the unemployment rates diverge widely: Maine's is 3.9%, while New Brunswick's is 9.1%.  Now what could account for that?  Globe and Mail business columnist Neil Reynolds discusses one study that came up with an answer.

They differ in one important way. Maine's unemployment insurance program is owned and operated by the state, which requires it to pay its own way. New Brunswick's is owned and operated by the federal government. Could the difference in ownership and management of these programs explain the extraordinary employment gap between these neighbours?

Well, yes, and now we have the academic research confirm [sic] it. Using Maine and New Brunswick as "a dynamic natural experiment," two economists have analyzed the unemployment history of these jurisdictions across 50 years. Chris Riddell (Queen's University) and Peter Kuhn (University of California at Santa Barbara) conclude in a report published earlier this year that Canada's notoriously generous UI benefits are indisputably responsible for New Brunswick's higher level of unemployment. The report is more proof that you get more of what you subsidize — including voluntary unemployment.

Before the 1950s, Maine had higher unemployment than New Brunswick.  After the Canadian federal government took over and expanded the unemployment insurance (UI) benefits program in 1953, unemployment in New Brunswick rose above that in Maine.  Then, when the feds again liberalised the requirements for receiving UI in the early 1970s, New Brunswick's unemployment rate increased still further.

[M]any Canadians came to regard UI as "Lotto 10-40," meaning you needed to work for only 10 weeks to collect UI payments for 40 weeks — and to do so without penalty year after year.

The Canadian UI program thus provides a large subsidy for part-year and seasonal employment that Maine's program does not.  (Maine has not altered its UI benefit rules significantly during this time frame.)

The researchers found the percentage of the labour force working six months or less of the year is two to three times greater in New Brunswick than in Maine.

Messrs. Riddell and Kuhn report this astonishing calculation of UI incentives: For every 10-per-cent increase in UI income for New Brunswickers who worked less than half the year, 10 per cent more people reduced their work to less than half the year.

The calculation is neat but, from an economics perspective, that finding is not surprising.  If you pay people for working only part of the year, you'll get more part-year workers.

Fortuitously, this study corroborates the point I made at the end of this post about differing unemployment rates in Canada and the US.

For access to the full column, click here.

Previous related post: Edmund Phelps wins Nobel for overturning economic myth