Priming the pump:
By Michael Smart and Kevin Milligan
The politics of federal spending in Canada
In the run-up to an election campaign, no interested observer of Canadian
politics could doubt that the government makes policy decisions with an eye to the poll
results. Government spending announcements in particular often seem designed to "prime the pump"
for government candidates in the next election.
This year is no different. Media sources have estimated the Martin government's spending
announcements since April 1 have totalled more than $5 billion. Indeed, it seems there is good
news for everyone this spring - from the fishing outports of the East Coast (unemployment
benefits) to the shipyards of British Columbia (naval procurement).
For some, this sort of chequebook politics must rankle: no one likes to be bribed with his or
her own money. But does it really matter? Most of the new announcements this year are in fact
of old money - funds that had been committed in previous budgets. While electoral calculus may
well be behind the media strategy of the Prime Minister's Office, there is less evidence that it
matters to the real spending priorities of the government.
Regional development programs: a case in point
To contribute to an understanding of these issues, we have worked to develop
hard evidence on the allocation of federal discretionary spending. We examined all grants
approved by the two largest federal regional development agencies for the 1988-2001 period and
calculated how they were allocated across federal ridings. The programs, known as Atlantic
Canada Opportunities Agency (ACOA) and Economic Development Agency of Canada for Quebec Regions
(EDAQ), give grants to businesses, non-governmental organizations and local governments for a wide
variety of capital projects.
Cash transfer programs like ACOA and EDAQ can serve important objectives of regional
development. Inevitably, however, the way in which subsidies are allocated among companies and
localities in the eligible regions is at the discretion of officials, and decisions on allocation
are rarely transparent. This raises the possibility that funding decisions reflect the political
calculus of the government of the day as much as legitimate economic development objectives.
Allegations of political interference in individual granting decisions have
been made about ACOA and the other regional development agencies. Norman Spector, who served as
president of ACOA in the early 1990s, has written of "dodging political entreaties to use my
agency as a slush fund." Of course, the existing evidence of political motivations in grant
allocations is anecdotal at best. In our work, we offer a more systematic look at the politics of
regional development grants.
Importantly, our data are comprehensive and cover a period during which two different political
parties were in power, and cover two regions (Quebec and the Atlantic provinces) with distinctly
different economic and demographic environments. Further details on the data set and the methods
we used to trace the grants to individual federal ridings can be
found on our
If politics are indeed driving the grant allocation, this could be manifested in a number of
ways. Grants might be used to reward political supporters, to "buy" votes in swing ridings, or
for a variety of other political objectives, as well as for their legitimate role as a tool of
regional industrial policy.
Courting the voter
As a starting point, we calculated the average level of spending per capita in each riding in
the five eligible provinces. Figure 1 shows the average per capita level of annual spending
commitments in ridings with opposition members of Parliament, government backbenchers, and cabinet
ministers. Ridings represented by government members received $65.86 (in year 2000 dollars) in
per capita spending each year, compared to $48.58 in opposition ridings.
Not all government MPs are equally successful in delivering support to their
ridings. Figure 1 shows that cabinet ministers' ridings received $78.88 per capita,
compared to $60.99 per person in government backbenchers' ridings.
Could these differences be a coincidence? Perhaps, but this is unlikely -
the differences are too large and too consistent for that. In other words, we can be reasonably
sure that the differences are not just a fluke.
Another possible objection to our results is that spending differences between ridings are
real, but they reflect other, purely economic influences on spending that happen to coincide with
our political variables. For example, an agency may be more likely to spend its funds in urban
ridings, and urban ridings may be more likely to vote for the government party. A more elaborate
statistical methodology that controls for factors like these, while too cumbersome to present
here, leads to the same qualitative conclusions as we draw from the simple comparisons of
Of course, political influence on spending patterns need not only result in greater spending in
government ridings. Instead, governments may tilt grant allocations in order to "buy" votes of
undecided voters in future elections. In Canada's electoral system, however, popular votes don't
really matter - only seats do. So a clever politician should allocate more grants to ridings
where the past margin of victory was small. These are the ridings where a few grant dollars are
most likely to "swing" the riding from the opposition to the government party.
To examine this possibility, we next split ridings into two equal groups,
based on whether the margin of victory in the preceding election was large or small.
Figure 2 shows the results. In ridings that had previously elected an opposition MP, the
half with the closest margin of victory again have higher spending - $61.89 per capita each year
for narrow-margin ridings, versus $35.18 for wide-margin ridings. For ridings that had elected a
government MP, however, the difference is negligible. Thus evidence supports the idea that funds
are allocated disproportionately to swing ridings that had previously gone to the opposition. But
on the government side, MPs with strong and weak local support fare equally well.
Last autumn, the minister responsible for ACOA faced questions in the House about the allocation
of spending to his own riding. Our data also allow us to examine this issue.
The ACOA ministry has been shuffled frequently: between 1988 and 2001, no fewer than eight MPs
held the portfolio. We can therefore examine how spending in each riding changes, relative to
other ridings, as the local MP is promoted to the ACOA portfolio. (By looking at the change in spending from baseline levels, we control for the other, non-political determinants of spending.)
Figure 3 shows average spending in ACOA ministers' ridings, relative to baseline
levels. In the year following an MP's promotion to the ACOA portfolio, per capita spending in the
local riding increases by $153.21 on average. Of course, the exact amount of increase differs
from case to case and time to time; the margins of error around the data points in the figure
indicate the normal range of variation above and below the average change. Notice that,
throughout this range, spending is significantly higher than before elevation to the ACOA
The effect is the same in the second year after appointment, though smaller: per capita
spending is $83.07 higher on average than before the appointment. In the third and subsequent
years (when the ACOA portfolio had generally been rotated to another member), spending remains
somewhat higher, but the difference from baseline becomes essentially negligible.
The statistics give an intriguing sense of political motivations in the
allocation of funds. There is evidence that regional development spending goes
disproportionately to government ridings and to "swing" opposition ridings, and that the ACOA
minister and other members of cabinet fare especially well. These findings seem troubling.
Regional grants can in principle serve a productive role in the Canadian economy, but that is
unlikely to be the case when political considerations rule over economic ones.
RELATED ITEM: Who helps politicians spend?
CBC does not endorse and is not responsible for the content of external
sites. Links will open in new window.
The authors, Michael Smart and Kevin Milligan, teach economics at the University of Toronto and the University of
British Columbia, respectively.