There is an old adage in Canadian political lore that “Tory times are tough times” and that Liberal governments have generally seen more prosperous times. Given that every Prime Minister since Confederation has been either a Liberal or a Conservative, it should be a relatively simple matter to find indicators rank the economic performance of each Prime Minister’s term to see what if any support there is for such a view.
Getting consistent data is no simple matter and getting indicators like unemployment rates or inflation before World War I is difficult. There are however, annual output and population estimates for the Canadian economy going back to 1870. As part of Urquhart and Green’s work on constructing GDP estimates for Canada from 1870 to 1926, Mac Urquhart put out a discussion paper on Canadian economic for growth from 1870 to 1980 (QED, Discussion Paper no. 734) that provided an estimate of GNP in current and 1981 dollars for the 1870 to 1985 period. Using this data for the period 1870 to 1960, Statistics Canada data on real GDP for the period 1961 to 2010, and annual population data for the period 1870 to 2010, I decided to construct the average annual growth rate of real per capita GNP/GDP for the term of each Canadian Prime Minister. Where the term is one year or less, the growth rate for that year is used which can be problematic given the brevity of some prime ministerial terms.
While the measure is simple, the results are interesting and are provided in the graph below, which simply ranks the terms of Prime Ministers from lowest to highest. Over the period 1870 to 2010, the average annual growth rate of real per capita GNP/GDP was 3.2 percent. There has been considerable variation across prime ministerial terms with some of the worst performances during Conservative prime ministers. In terms of above average performers, the Prime Ministers are Mackenzie-King (1935-48), Borden (1911-20), St. Laurent (1948-57), MacDonald (1871-73), Laurier (1896-1911), Turner (1984), Pearson (1963-68) and Diefenbaker (1957-63). Everyone else is below the average of 3.2 percent. When it comes to above average performers, three were Conservatives and five are Liberal though given the brevity of Turner’s term it can probably be removed. As for the nineteen below average performers, twelve were Conservatives and seven were Liberals.
The only conclusion I draw from all this is not that Liberal Prime ministers have generated higher per capita income growth but that correlation is not causation. For example, Laurier and St. Laurent both preside over periods of economic boom – the Wheat Boom and the Post-War boom while Mackenzie-King (1935-48) has a term that includes a world war. Put a Conservative in power during a boom period or a war and they do well also – Borden for example. The Conservatives do appear to have had a knack for being in power during periods of severe economic distress – particularly the Great Depression of the 1930s as well as the severe recessions of 1920-21 and 1926 – and therefore are weighed down by the performance of Meighen and Bennett. When Liberal prime ministers are caught smack in the middle of a depression, they also show declines in real per capita output as in the case of Alexander McKenzie who was prime minister from 1873 to 1878, an era that until the 1930s was also viewed as a Great Depression. As well, Mackenzie-King’s second term from 1926-30 coincides with the debut of the Great Depression and is also marked by negative per capita income growth. Had he remained in power for the 1930 to 1935 period, it is likely the growth of the World War II period would have been associated with a Conservative Prime Minister. In politics, timing is everything.
Figure: Prime Ministers Ranked by Average Annual Real Per Capita GNP/GDP Growth
Worthwhile Canadian Initiative 
My colleagues Steve Ferris and Marcel Voia have done some interesting work on this – they find that the coefficient on “liberal prime minister” is positive and significant in a regression explaining Canadian economic growth over time. Part of the reason seems to be more expansionary monetary policy.
Instead of simply asking “How was the Canadian economy at this time?” wouldn’t it be better to ask “How was it relative to other countries?”. Preferably compared to similar countries.
How does this outcome change when we also control for some economic output factor like change in GDP? Frances, what other factors does Ferris’ paper look into? I havent had time to read it yet but plan on doing it at some point after my classes are over.
So seeing as how Mackenzie King is on both sides of the scale, we can safely assume it doesn’t matter who is in charge.
I think Mackenzie-King nails it as being irrelevant.
He wasn’t that much of a social spender before 1945 either. Bennett had proposed a “New Deal” of his own in the dying days of his government which gave us the Bank of Canada and the CBC. Much of it was ruled unconstitutional by the Supreme Court and the Privy Council. Mackenzie-King just sort of sailed along until 1939 when the storm clouds gathered.
It wasn’t until 1944 after the Beveridge Report in the UK and the NDP winning in Saskatchewan that the Liberals took a serious left-turn and became the left-wing part we know from the post-war years.
I think the causality runs the other way, but that there may be something there (the US has a similar story, although it goes the other way in Britain). Here is my back of the napkin story:
I think commodity booms may tell part of the story. There are two Canada’s – an industrial core and a resource extracting periphery. Our political parties reflect these cleavages – the Conservatives are the party of the periphery (although historically they were the party of the core – I’d peg the switch as occurring postwar), while the Liberals are the party of the core.
Because the population of the core is bigger, in general the core party dominates Canadian politics. However, every now and then there are sustained commodity booms. While commodity booms drive up the growth rate, they also cause a relative shift of power and influence to the periphery, enabling the periphery party to take power. Often, because the shift in power lags the commodity boom, they tend to become dominant around the peak of the boom (though sometimes the boom lasted longer).
It works for most of the interregnums of rule by the core party:
-Laurier (gold rush)
-King (WWI commodity shocks)
-Diefenbaker (post-Suez oil spike)
-Clark/Mulroney (oil spike, other commodities rising in prices)
-Harper (2000s oil spike, rising food prices)
Interesting, but I suspect it would be more telling to rank the PMs by the performance of the economy in the twelve months leading up to election day, and whether or not the outcome was change of government or preservation. That way you capture public mood about the economy, and put less emphasis on who’s in charge.