Political Sabermetrics Part 1: The Big Picture

Last weekend, I mentioned that I'll be taking a look at how Canadian political parties spend their money to see what role money plays in our political scene, and how parties can make better use of it. While most of that discussion will take place at the riding level, it's worth starting off with a look at the national scene to see both how money was spent in the 2006 election, and what kind of return parties received on it.

Let's start with a table of what strikes me as the most relevant data from the election, followed by some discussion of what it means. First off, the basic data from the 2006 election:

Party	Election Expenses	$ Reimbursed	Votes
Conservative Party of Canada	$18,019,179.28	$9,009,589.64	5,374,071
Liberal Party of Canada	$17,447,130.00	$8,719,845.00	4,479,415
New Democratic Party	$13,524,524.81	$6,735,433.46	2,589,597
Bloc Québécois	$4,527,629.97	$2,261,702.49	1,553,201
Green Party of Canada	$910,979.08	$455,489.54	665,940
Christian Heritage Party	$63,257.45	n/a	28,152
Progressive Canadian Party	$5,777.38	n/a	14,446

By way of explanation, the second column features each party's election expenses as declared in its Election Expenses Return (see row 10 of the PDFs at the link). For parties who receive a sufficient amount of a riding or national vote, the amount in the third column represents a roughly 50% reimbursement of its election expenses. To my knowledge this funding isn't yet available to smaller parties, unlike the $1.75/year/vote annual funding which now covers all registered parties.

Let's add in one more chart which helps to interpret the above numbers:

Party	Annual Funding	$/Vote
Conservative Party of Canada	$9,404,624.25	$3.35
Liberal Party of Canada	$7,838,976.25	$3.89
New Democratic Party	$4,531,794.75	$5.22
Bloc Québécois	$2,718,101.75	$2.92
Green Party of Canada	$1,165,395.00	$1.37
Christian Heritage Party	$49,266.00	$2.25
Progressive Canadian Party	$25,280.50	$0.40

The amount in the second column is determined by multiplying each party's vote total by $1.75. Keep in mind that while all of the other numbers in the table are of one-time application, this one determines federal party funding on an annual basis until the next federal election.

Finally, the third column reflects each party's 2006 election expenses divided by the number of votes received. Note that while parties will generally want to be efficient in allocating their money, that doesn't mean that a low number necessarily reflects good planning: indeed, it may be highly inefficient for a party to choose to spend a low amount of money if it could have won more votes (and resulting funding) by spending more.

Note that I've included a couple of extra parties in the vote standings beyond the ones which contested all ridings nationally or regionally. My purpose in doing so is to provide some indication of how effective funding is for various party sizes...and I suspect I'll end up going back to add more parties to examine those implications in more detail.

So what can we learn from this chart? Let's stick with a couple of general observations for now.

First, the election funding system appears to offer enough money back to make elections at worst cost-neutral for Canadian political parties. Based on the combination of election-spending reimbursements and per-vote funding, every party listed would more than make up its election investment within two years of an election - and some within a matter of months.

Of course, parties would still want to pick up as many donations as possible for non-election spending. But while individual donations still have some significance, a party which spends money well during elections can bring enough money in to get by even if its outside fund-raising is unimpressive.

Second, let's note that there appear to be several possible strategic niches which a Canadian political party may pursue. It's possible for a smaller party to build itself up by spending little money and amassing resources over time. At very low levels of spending, it looks like virtually any additional financial input will turn out to be money well spent based on the per-vote funding. And at a relatively moderate spending level, a party can boost its financial returns again if it reaches the reimbursement thresholds (5% of the vote within a riding or 2% nationally).

In contrast, the cost appears to be significantly higher for the national parties which seriously contest seats. Any party looking to put itself in play for spots in Parliament can expect to spend a lot more money per vote in doing so - and indeed will likely have to spend at or near the overall limits (for a riding at least, and arguably nationally as well except for the Bloc) rather than seeking out the best possible financial return on investment.

I'll leave the discussion there for now. But there'll be lots more worth analyzing from both a systemic and a party level later on.

(Edit: split tables based on formatting.)