Canadian governments tend to raise spending in elections years.
True story: In 1980-something I ran for office, campaigning for class president in the 5th grade.
The competition was tough. My opponents had big ideas like less homework and longer recess.
My platform lacked ambition. It wasn’t transformational; it was merely transactional: If elected, I’d host a class party and bring the pop and chips.
This strategy had several strengths:
- If I won, the electorate felt they’d get something for nothing;
- They’d get it soon, not far into the future;
- My “party platform” was easy to explain and remember; and
- It was also credible because voters believed that I’d deliver on my one promise.
I won the election and hosted that party (wearing my best Beaver Canoe track suit!). Afterwards, things returned to normal as being class president didn’t involve any actual duties.
Looking back now, and with a big election underway, I wondered was my short political career as a kid an aberration, or was it part of a larger electoral strategy used by adults?
Some evidence on whether Canadian governments try to buy votes
Using past research in political business cycles as a guide, let’s investigate this issue with annual data on elections as well as economic and budget outcomes for federal and provincial governments in Canada for 1990 to 2013.
If trying to buy votes is common, then budget balances should tend to be worse in election years, when governments raise spending and/or cut taxes to woo voters.
The table below shows results from fixed effects panel regressions, where I try to control for economic factors that might affect fiscal performance (as well as general differences across jurisdictions, provided they don’t change over the sample).
As predicted, the shaded row shows that in election years the budget balance is about 0.4 percentage points of GDP worse on average, holding these other factors constant (column 1).
Most of this difference is due to increased program spending (column 2) — responsible for about 70% of the deterioration in the government’s fiscal position — with a smaller contribution from lower revenues, likely due to tax cuts (column 3).
These results don’t suggest that right-wing parties had significantly better budget balances, but they did have smaller governments with less spending and less tax revenue collected as a share of the economy.
Our newsletter about the public service.
Nominated for a Digital Publishing Award.
I’m not surprised by these results and I don’t expect you to be either. If anything, I’m only surprised at how robust this “election year effect” seems to be to different regression specifications (other economic variables, sub-periods and time fixed effects, not reported here).
Now there could be other explanations. Budget balances tend to be better in the year before the election and worse during the election. So, it’s possible that governments bunch their spending closer to elections so they can make big program announcements that people may remember when they vote (or they may call elections when spending is elevated).
Even if these attempts at what I call “vote buying” are really what’s happening, this evidence doesn’t mean that all governments did this — though it seems that many did. Furthermore, the analysis stops in 2013, so it doesn’t include the last few years.
Also, I only looked at what I thought were attempts to win votes, but I didn’t analyze whether it tends to “work”. It’s hard to say whether this actually boosts incumbents’ re-election chances, or whether there were any impacts on the economy.
So what can we, as voters, do about this?
First, we should see these efforts for what they are and resist them, where appropriate. We don’t really get “something for nothing”. Not from the government or from opposition party promises. Free lunches are hard to come by because there are economic costs to raising government revenues that fund higher spending.
If our votes weren’t seen as being for sale, then politicians might stop trying to bribe us… with our money.
 Namely, the jurisdiction’s nominal GDP growth, the change in its unemployment rate, the level of the Bank of Canada’s monetary policy interest rate and the change in the Bank of Canada’s commodity price index.
 Studying each jurisdiction separately, this vote buying effect appears to have been largest in Alberta (from tax cuts) and PEI (from increased spending).