Monday, May 17, 2010

Money and Canadian Votes

Elections are expensive things. Altogether, the five major parties in Canada spent over $58,000,000 on the 2008 election. But just how valuable was that spending? Which party got the most for every dollar spent, and what did they spend it on?

That's the topic for today's post. Below, you'll see a chart comparing the amount of money each party spent in order to receive one vote. While I do intend to look at the Bloc Québécois and the Green Party, their special circumstances make them less useful as a basis for comparison. The three major parties, however, are on the same playing field: they are well-known parties, have long standing organizations, and run in every part of the country.Clearly, from this it is obvious that the Greens have gotten the best bang for the buck. But as they start to spend more money ($498,000 in 2004, $911,000 in 2006, and $2,796,000 in 2008), the amount spent on each voter goes up. However, despite this, they are still the best performing party with only $2.98 spent per vote earned in the 2008 election.

The Bloc, historically, has been the second-best performing party. This should come as no surprise, as they run in one province and so everything they do is magnified and they are in the spotlight more often. They've become less efficient with their use of money, however, going from $2.68 per vote in 2004 to $3.53 per vote in 2008. The lower voter turnout, however, plays a role in this. They have spent the same amount of money in each of the elections: $4,502,000 in 2004, $4,544,000 in 2006, and $4,876,000 in 2008.

But what about the other parties? Aside from their electoral loss in 2004, the Conservatives have made the best use of their money in the 2006 and 2008 election - and this with only marginal increases in spending each year, going from $17.2 million in 2004 to $18.0 million in 2006 and $19.4 million in 2008. Their best result was in the 2006 election, when only $3.35 was spent per vote, compared to $4.30 per vote in 2004 and $3.73 per vote in 2008.

The Liberals have become steadily worse in their use of their own money, rising from $3.33 per vote in 2004 to $4.00 per vote in 2008. And that was with less money spent in 2008: $14.5 million as compared to $17.4 million in 2006 and $16.6 million in 2004.

Finally, the New Democrats have always spent more per vote than the other parties, with $6.66/vote being spent in 2008, compared to $5.22/vote in 2006 and $5.65/vote in 2004. They have also increased the amount of money they have spent, with $12 million in 2004 rising to $13.5 million in 2006 and $16.8 million in 2008. Yes, in the last election, the NDP out-spent the Liberals.

But what does this mean? Out of the three major parties, the Conservatives have gotten the best bang for their buck lately. In the last two elections when they have out-performed the Liberals and the NDP, they have spent less of their money on their leader's tour and party salaries than the other two. In 2006 and 2008, the percentage of total money spent on the leader's tour was 17% and 13%, respectively. The Liberal numbers for those two elections are 19% and 17% respectively, and the NDP's are 27% and 24%, respectively.

This seems to indicate that flying your party leader around the country is not an efficient use of your money, as opposed to, say, advertising. Of the three parties, the NDP spends the least amount on advertising and has the worst results.

In the last two years, the Conservatives have spent about 4-5% of their total on salaries, compared to 10% for the Liberals and NDP. It seems that having a large group of paid employees does not help in getting votes.

When we take the last three elections and average them out, we get $3.79/vote spent for the Conservatives, $3.74/vote for the Liberals, $5.84/vote for the NDP, $3.05/vote for the Bloc, and $1.74/vote for the Greens.

From this, it seems to suggest that for the main, traditional parties, the amount of money spent can dictate the amount of votes earned. The Conservatives and Liberals have averaged about the same $/vote since the Conservatives were re-born in 2003. This means the Tories are at an advantage because they currently have more money.

The NDP has to spent a lot more to get votes, and it doesn't appear that spending more means getting more votes. Spending has out-paced vote growth for the NDP, indicating that the fact the NDP spent like the "big boys" in 2008 had less to do with their electoral success than the performance of the party itself.

Over the last three elections, the NDP has averaged 46% of its expenses spent on advertising. For the Conservatives, that number is 49% and for the Liberals it is 57%. Perhaps the NDP needs to spend more on advertising - their best electoral result, in 2008, was the time they spent the most on advertising.

The Bloc has averaged 54% over that period on advertising and the Greens 38%. That last number is misleading, as they spent 65% of their money on advertising in 2008 but only 15% in 2006. And 2008 was their best electoral result.

When it comes to party salaries, the NDP has spent 10% of their expenses on that, compared to 4% for the Conservatives and 8% for the Liberals. The Bloc has averaged 12% and the Greens 32% (again, misleading, as they spent 48% of their money on salaries in 2006 but only 4% in 2008).

It is difficult to come to some solid conclusions from these numbers. The Bloc and the Greens are better at using their money, but that may be caused more by their special circumstances. The Liberals have gotten worse, but on the whole are not too different from the Tories. The NDP spends way too much per voter.

In 2008, when of the three major parties the Tories performed best in $/vote, they spent the most on advertising (55%) and the least on the leader's tour (13%) and salaries (4%). The NDP performed worst, and spent the least on advertising (50%), the most on the tour (24%), and the most on salaries (10%).

In 2006, when the Tories again out-performed the other two parties, they spent the least on the leader's tour (17%) and salaries (5%), but were middle-of-the-pack on advertising (51%). The NDP performed worst again in this election, and spent the least on advertising (44%), the most on the tour (27%), and tied the Liberals with 10% spent on salaries.

Finally, in 2004, when the Liberals performed the best, they spent the most on advertising (61%), but were middle-of-the-pack on the leader's tour (17%) and salaries (6%). The Tories did worse than the Liberals in this election, and spent less on advertising (42%), more on the tour (22%), and less on salaries (3%). The NDP spent 45% on advertising, 3% on the tour, and 10% on salaries.

The 2004 result muddies the water a little, but it seems that when a party spends more money on advertising and less on an expensive tour and salaried employees, they perform better.

11 comments:

  1. Do I understand correctly that you are considering only the election expenses reported by the national parties during the campaign period in this analysis? I.e., you're excluding candidate expenditures in the ridings, I believe.

    Also, when you say "best bang for the buck" I take it you mean in terms of votes received, rather than seats won?

    It would be difficult to draw firm conclusions without a dataset that encompassed more elections. Also, when you say that one tour is more expensive than another, that might indicate that the party's target seats are far away from one another. On the other hand, that tour might generate a lot of earned media, whose value might be worth double that of paid advertising.

    Another factor to consider is whether there is a lag time across elections. Of necessity, the economics are going to be different for parties having momentum, and those experiencing the declining part of a 20-year cycle. Perhaps in order to grow in one campaign, one had to invest in running a previous campaign more professionally.

    Anyways, a few points worth considering.

    ReplyDelete
  2. but it seems that when a party spends more money on advertising and less on an expensive tour and salaried employees, they perform better.

    Which actually makes logical sense. More people see advertising than see either a leader or in many cases the local candidate.


    Now what would be interesting Eric, if it's available, is the proportion spent on media?

    i:e: how much for TV, for radio for national papers, for local papers and flyers etc?

    ReplyDelete
  3. --- Do I understand correctly that you are considering only the election expenses reported by the national parties during the campaign period in this analysis? I.e., you're excluding candidate expenditures in the ridings, I believe.

    I used the numbers here: http://www.elections.ca/content.asp?section=fin&dir=pol&document=table2_08&lang=e&textonly=false

    --- Also, when you say "best bang for the buck" I take it you mean in terms of votes received, rather than seats won?

    Yes, indeed.

    --- It would be difficult to draw firm conclusions without a dataset that encompassed more elections.

    I agree, but we've only had three elections with these parties. When you go back to 2000 or before, it becomes harder to draw conclusions that can be compared to today's situation.

    --- i:e: how much for TV, for radio for national papers, for local papers and flyers etc?

    Elections Canada breaks it down into Radio/TV and "Other".

    ReplyDelete
  4. Elections Canada breaks it down into Radio/TV and "Other".

    Darn. Would really like a better breakdown than that.

    Bang for the buck"" if you know what I mean.

    ReplyDelete
  5. Right. So those are the national expenditures ("for a Registered Party"). The combined version (national + candidate) will be onerous to produce, but pretty soon I'll have the data to do it.

    I take your point about the current configuration of parties. That also implies the current state of the electoral playing field. As Nate Silver noted in relation to the literature in his recent work on the UK election, there are normal/incremental elections and realigning elections.

    The 3 you use show the Conservatives on an upswing (and the NDP as well, to a lesser extent) and the Liberals struggling. Will the next election be an incremental change one way or the other, or will it see the kind of complete upheaval we saw in 1993? I think until we run some correlations across an entire electoral cycle, we won't have the definitive answers.

    However, there's an article just accepted for publication in the Journal of Elections using data from Australia that you might find interesting in this vein:

    Coram, Alex, "Resource spending over time in competitions for electoral support", Journal of Elections, Article in Press, Accepted Manuscript, May 12, 2010

    Cheers.

    ReplyDelete
  6. Is anyone at all surprised that the socialists spend more to produce less?

    I do wonder if seats won wouldn't make more sense as a metric, though. The objective is to win seats. The votes themselves have no intrinsic value.

    ReplyDelete
  7. This analysis is useful, but a conclusion that advertising is more effective than a leader's tour should also take into account the size of the campaign budget.
    The most effective political ads are the most expensive, prime time ads. Only the parties with the biggest budgets can afford them. The NDP could spend a bigger percentage of its campaign budget on ads but it might not necessarily help if all they can afford are a very tiny number of prime time ads, or a bunch of late night ads that nobody sees. The leader's tour is not a bad strategy with a tight budget because the TV coverage is free.

    ReplyDelete
  8. There is a huge missing variable in all this, and that is the free press politicians get from the media. In 2004, as a new Prime Minister, Martin was able to dominate in terms of the press. In 2006, due to his strategy of early morning policy announcements, it was Harper who dominated the headlines. Harper similarly dominated in 2008.

    The NDP cannot get the same kind of coverage that the big parties get, but does have the financial wherewithal to spend the maximum during elections.

    The Bloc and Greens get a disproportionate amount of attention relative to their size. Both were in the debates, and the Bloc is prominent in French election coverage. The Greens benefit from the fact that a fair proportion of the "chattering classes" are Greens.

    ReplyDelete
  9. Heh. Greens getting more media attention than our votes justify. Hehehehe. If only.

    Between elections we are lucky to get a single mention in a paper. Google news search...
    "Elizabeth May": 59
    "Gilles Duceppe": 122
    "Jack Layton": 221
    "Michael Ignatieff": 464
    "Stephen Harper": 3,528

    Wow, even more of a spread than I thought.

    Yes, during the election we get more coverage but I suspect any measure would put it well under 10% of total coverage, and probably under 1 minute on TV for every 10 for any other leader (outside of Duceppe).

    Now, Harper getting more coverage at the moment makes a lot of sense since he is the prime minister but why does Layton get nearly 4 times the coverage? Ignatieff nearly 8 times? Current polls don't put that big a spread in place.

    As to the elephant in the room for spending, namely the local spending, it is a safe bet the CPC & Liberals outshoot the NDP & Greens by a mile. My local riding had the Green vote just barely under 10%, over the NDP. Checking Elections Canada you can see the Greens spent $1500, NDP $800, CPC $67k, Liberals $71k. Now _there_ is a spread.

    Guelph was a very tight 4 way race (or at least was viewed as one) and the spending there ...
    NDP: $60k
    CPC: $63k
    Green: $77k
    Liberals: $90k (winners)
    Heck, even the Animal Alliance Environment Voters Party of Canada got into the act by spending $5k.

    That race was even worse thanks to Harper breaking his own law - there was a by-election where the NDP/Liberal/Greens all blew over $80k each and the CPC $63k and it was canceled at the last second for the general election.

    Yes, that by-election being cancelled really hurt the Greens as we didn't have the $80k to blow unlike the Liberals/CPC/NDP and a double length election campaign caused no shortage of burnout.

    Back to the topic though. In ridings where your party is in contention you will likely spend $50k plus. Where they are not (and the NDP/Greens have a lot of those) you won't. Mix that in and you'll see even more drastic results I'm sure.

    The data is from...
    http://www.elections.ca/scripts/webpep/fin2/select_election.aspx?entity=1&lang=e

    ReplyDelete
  10. Votes do have a value, around $1.95 per year to the party. Thus a party that spends like the NDP did needs almost 4 years to recover their costs directly (although the massive 50% return for party expenses would cut that down to 2 years).

    ReplyDelete
  11. I like the breakdowns on salary vs. advertising, but it would be even more interesting to see this broken down in positive vs. negative advertising as well. This would be a little difficult, granted, but I think it would go a long way towards assessing the efficacy of advertising expenditures by political parties (in benefit to a party vs. damage to alternative parties.)

    All this is by way of saying the following: I might like to look into this, and I was wondering if you would be able to tell me whether your sources for these figures were restricted to the parties various financial disclosures, and if not, whether it might be possible to obtain any additional information for purposes of the above comparison.

    Best,

    Andrew

    ReplyDelete

COMMENT MODERATION POLICY - Please be respectful when commenting. If choosing to remain anonymous, please sign your comment with some sort of pseudonym to avoid confusion. Please do not use any derogatory terms for fellow commenters, parties, or politicians. Inflammatory and overly partisan comments will not be posted. PLEASE KEEP DISCUSSION ON TOPIC.