Friday, October 30, 2009

Someone teach the National Post basic business journalism please

I’m a business journalist in my day job, and while today my reportage is more technology and marketing focused, back in the day I used to have to do a lot of spot reporting on company earnings. Wade through a mess of confusing numbers and acronyms – EBIDTA, GAAP, one-time charges, revenue, income, currency charges – and quickly discover the facts, what’s newsworthy and why, and report it.

And I tell you, if I got it as wrong as the Post did here, I’d be fearing for my job. Their analysis/link they try to make is completely off base and without merit, and is completely unsupported by the facts. With the Post’s death rumoured in recent days I’ve been defending them; not because they occasionally publish my commentaries but because I think we need more media organizations in this country, not less, because as a journalist I support jobs for journalists, and because off the op/ed pages, generally the Post is no more or less biased than any other outlet. But pieces that are completely full of crap like this one make it difficult to do so.

First of all, they rely on a Q2 2009 to Q3 2009 comparison as the main plank to support their thesis: dropping Liberal polling equals dropping Liberal donations. Ask any business journalist, financial analyst or accountant and they’ll tell you comparing consecutive quarters is largely meaningless. That’s because business is seasonal, it’s cyclical. So comparing, say, a department store’s Q4 vs Q1 next year revenue will likely show a big drop always. Why? Because Christmas is in Q4, and it’s a major shopping season. Sales always taper off in January.

That’s why comparing consecutive quarters is meaningless, and it’s why instead it’s standard practice to compare to the year-ago quarter. So, comparing Q4 this year versus Q4 last year and you’ll be comparing two like shopping seasons, giving you a better trend line to judge performance.

Second, if you want a true apples to apples comparison you need to look for and exclude one-time expenses or revenues. A company might take a big hit in one quarter because of a fine, for example. Or realize a big one-time gain because of an asset sale. Comparing numbers without factoring-out one-time numbers gives you a meaningless conclusion.

Anyway, here’s the Q2 to Q3 numbers as Tweeted by Canwest’s David Akin:

Q309/Q209 #cdnpoli Contribution $ only: #BQ -50% #CPC -5% #GPC +14% #LPC -50% #NDP +57%
As I said, comparing consecutive quarters is largely invalid, and there’s one major factor the Post missed in drawing the conclusion they did from these stats (I don’t quarrel with Akin here, he’s just reporting the numbers): the Liberals had a huge one-time revenue boost in the previous quarter: the Vancouver convention. Over 2000 people at up to $1000 bucks a pop. That skews those numbers.

Of course, they dropped from that quarter’s result, and it had little to nothing to do with polling: it had everything to do with not having a major national convention every quarter. To not factor that in is misleading, dishonest and just plain crappy reporting.

That aside, let’s look at the Post’s wider conclusion: that Liberal polling decline = Liberal donation drop. That’s only supportable if you ignore the methodology flaws already noted and look at just the LPC numbers in isolation. But their theory doesn’t hold with the other numbers. The Cons shot-up in the polls, but their contribution numbers dropped 5%. The NDP tread water in the polls at best, and they went up 57% in contributions. Their thesis just doesn’t make sense, even if you ignore the other obvious flaws in their supporting data.

Now, all that said, what happens when you do the comparison that is generally accepted practice in business, and compare to the year-ago quarter? Here is Akin’s tweet for that number:
Q309/Q308 #cdnpoli Contribution $ only: BQ -92% CPC -41% GPC -67% LPC +5% NDP -41%
Well, that’s a very different number isn’t it? Remember, we were fighting an election in this quarter a year ago, usually good for a nice fundraising boost. So it’s unsurprising the Conservatives and NDP are both off by 41%, year over year. Again, one-time rev boost in the year-ago quarter.

Nonetheless, that rev boost this time was available to all the parties, not just one (unlike the LPC convention) and yet only one party is not down sharply year over year. Indeed, only one party is actually up year-over-year for Q3, despite not having an election boost this year: and it’s the Liberal Party, up 5%.

So that just takes the Post’s theory and proves it to be completely full of crap. The LPC has fundraising momentum. There's a long ways to go to match the Conservatives in sheer volume, but the trends remain positive.

Now, could poor polling make it difficult to raise money? Absolutely. It will definitely make it harder, and things need to turn around on the polling front. But as much as the Post may wish to twist the laws of accounting, basic business sense and, well, reality to make it so, there’s nothing in the donation figures to show that has happened yet.

P.S. Visit Pundit Guide for a more non-partisan take.

Recommend this Post on Progressive Bloggers


ottlib said...

I disagree with you Jeff. The sooner the National Post goes the way of the Dodo the better off Canadian democracy will be.

For a democracy to work well it needs media outlets that will report facts and hard news not be the in-house propaganda tool for a political party.

The National Post does not even try to be objective but tries to claim that it is. At least the Sun Group virtually admit they are Conservative newspapers. The Post does not.

Its death cannot come soon enough for me and to hell with all of those journalists who sold out the princples they learned in journalism school for a job. They made their bed they can lie in it.

The Pundits' Guide said...

Hi Jeff, and thanks for the referral.

There's probably one last point to make, to provide the most complete picture possible, and that's the validity of the comparison to even last year's third quarter.

As you're aware 2008-Q3 was not a strong one for your party at all, given that an election was going on at the time. Taking it as the starting point for comparison could have made almost any amount raised this year look big ... and that's no more valid than what the (probably snarky) one off quickie blogpost at Full Comment tried to claim. [I'm prepared to cut them a bit of slack today ... I'm sure everyone was sitting on tenterhooks, and lord knows the unemployment rolls don't need to get any longer.]

When I first looked at contributions by date last year, in response to Hélène Buzzetti's story on the "Ignatieff effect", I noticed that Liberal fundraising during the campaign had dropped down to almost nothing. The Liberals and Conservatives have a different pattern of fundraising from the smaller parties, in that they tend to raise money ahead of campaigns, and then do very little during the election. The smaller parties are able to use the excitement of the campaign (and their increased media coverage) to generate the momentum necessary to raise funds, and this pattern and difference holds up at both the national and local levels. (I also later learned that Elections Canada ruled weeks before the campaign started that fundraising costs would henceforth be counted as election expenses, a significant change that may have had its own impact). And, even for an election campaign, that amount was small by Liberal standards.

So, I knew 2008-Q3 was unusually low, and thus didn't really look at it when making my assessment of the Liberals' performance this year. Rather, last month I had set out the criteria I would be looking to, for both the Liberals and NDP, and argued that raising something along the lines of $2M to $2.5M (about half of what had already been raised in the first half of the year) would signal that the Liberals were maintaining their new fundraising trend.

Today's numbers were in that ballpark, and also placed the party close to their entire fundraising totals of 2005 at the end of the third quarter of 2009, and well within striking range of its 2006 annual totals.

Thus I do think that saying the Liberals are the only party to increase year-over-year is juicing the story a tiny bit ... but of course I fully understand the desire to do so.

Again, thanks for the link, and stay tuned for the fine-tuned analysis by contribution size and date coming within a day or so.

Jim Parrett said...

This is totally wrong. The National Post is part of a monopoly of right-wing journalism in Canada. Its death will mean better things for Canada. It will mean perhaps a more-balanced media for Canadians - if it doesn't get bought up by another right-winger. Publishing the very occasional article by a progressive does not make the majority of NP's Murdoch-like propaganda on behalf of the Cons right. That tiny representation of the left in NP makes gives the editors all the more ammunition to load the gun on their side of the argument.