Lyle Stafford
Frances Woolley is a professor of economics at Carleton University
In Canada, the prime minister chooses when to hold an election. If a strong economy increases the government's chance of re-election, prime ministers will tend to call elections during good economic times.
Also, the governing party has considerable control over monetary and fiscal policy. If hard choices and painful budget cuts are necessary, a government is well advised to do them early in its mandate. By the time the next election rolls around, voters will have forgotten about the earlier pain. Their support can be regained if growth is strong and jobs are plentiful.
I've always believed in the political business cycle. How could politicians possibly resist the temptation to manipulate the economy and the timing of elections for their own benefit? But generally studies have found little evidence of such a cycle in Canada.
Now a paper in the Canadian Journal of Economics by my Carleton University colleagues, Stephen Ferris and Marcel-Cristian Voia, has found new evidence that elections affect economic growth.
Ferris and Voia do two things differently from previous studies. First, they look at more elections - every federal election from 1867 to 2005. Second, instead of looking at actual election dates, they consider when an election is likely to be held.
Economic management takes time. A government wanting a robust economy on election day has to start steering policy in the right direction at least a year beforehand. But no government knows exactly when an election will occur. Actual election dates are influenced by poll numbers, the weather, and so on. Yet a government can guess how likely it is to call an election in, say, 2012, based on factors such as the size of its majority and the length of time it has been in office.
Ferris and Voia's insight is that economic and fiscal policy will change course during the lead up to the expected election date. So they use political and economic variables to predict, for every year between 1867 and 2005, the probability of an election being called. For example, in 1994, with the Liberal government just one year into a strong majority mandate, the probability of an election was low, whereas in 1925, when MacKenzie King's government was in the fourth year of a minority mandate, the probability of an election was high.
Ferris and Voia find that, in the years since 1945, a 10 percentage point increase in the probability of an election being called increases GDP growth by 0.29 percentage points, controlling for all other factors - for example, from 2.0 to 2.29 per cent. Their interpretation: "political opportunism."
But other factors also affect economic growth.
Taking all the years from Confederation to 2005, Ferris and Voia find that economic growth is 1.61 percentage points higher under Liberal governments than under governments of any other political persuasion.
Follow Economy Lab on twitter