A number of American media commentators have recently taken aim at the fallacy of the “I’m not a scientist” argument. “I’m not a scientist” is an increasingly popular statement from American politicians who don’t believe in climate change. Whenever these politicians are presented with evidence that suggests climate change is real, they say “I’m not a scientist”, and think that excuses them from commenting on the evidence that contradicts their position. But, as several commentators have pointed out, it’s not acceptable for politicians who make legislative decisions on climate change to not be informed about it – and they don’t have to be scientists to do that. Politicians don’t have to be experts on everything, and they shouldn’t be expected to, but they do have the responsibility to know something about the issues they vote on.

North of the border, we in Canada now seem to have our own version of the “I’m not a scientist” argument. It’s the “I’m not an expert” reasoning. The “I’m not an expert” reasoning tends to arise whenever a member of the Conservative federal government uses questionable information, and then claims that the information must be accurate because it came from “experts”.

A few weeks ago, Finance Minister Joe Oliver used this reasoning while testifying at Canada’s House of Commons Finance Committee. At the Finance Committee meeting, he was asked about an estimate that a proposed reduction to Employment Insurance (EI) premiums for employers would create 25,000 jobs. That 25,000-jobs estimate came from the Canadian Federation of Small Business (CFIB). However, there was another estimate that had also been produced. This estimate came from the House of Commons’ own Parliamentary Budget Office (PBO). The PBO estimated that a net total of 800 jobs would be created by the EI premium reduction – and also estimated that 10,000 jobs could be lost because of other changes to EI included in the same legislation.

The PBO’s estimate is in this report. The report explains that the 800-job estimate is “based on the macro‐economic increase in employment resulting from an increase in aggregate demand and supply as a result of higher after‐tax business incomes (that is, the dollar increase in real GDP of leaving an extra dollar of EI revenues in the hands of business owners)” (p. 5-6). The report also notes that the 800 jobs would be created over only two years and that “over a longer outlook, there are few, if any, net employment effects from this measure” (p. 6).

Here and here are the CFIB’s two press releases on its estimate. Neither document contains any explanation of how the 25,000-job number was calculated, other than stating that it is an “estimate”.

Several members of the House of Commons Finance Committee asked Oliver to explain how the CFIB arrived at its estimate –  and to explain why the government was relying on the CFIB’s numbers, rather than on the PBO’s numbers. Here’s a summary of Oliver’s response:

I am aware that [the CFIB] have spoken to their members and they do the regular type of analysis that you’d expect them to do….[W]e have received a number of estimates from a variety of organizations on a variety of topics. Sometimes one looks at them and decides on the face whether they seem to make sense. When the CFIB says it’s a big, big deal for small business and it’s good news for people looking for jobs, we’re influenced by that…because they have the expertise on the ground. We are satisfied with their analysis…[W]e didn’t think we needed to do another analysis when we already had received one.

It’s embarrassing that a federal cabinet minister would use the “I’m not an expert” reasoning to promote an estimate with no explanation behind it  – and to ignore an estimate which was also created by experts, with clearly described methodology.  Also, the PBO’s work is funded by the federal government, through the government’s budget for the Library of Parliament. So the PBO report was indirectly paid for by the same government that then ignored the report – and this is a government obsessed with reducing allegedly wasteful spending.

Aaron Wherry, a reporter for Macleans magazine, was also puzzled by why Oliver was endorsing the CFIB’s estimate instead of the PBO’s. He contacted the CFIB to ask how it came up with the 25,000-job figure. The CFIB’s Dan Kelly told him that the proposal to reduce employers’ EI premiums was initiated by the CFIB, through “back and forth discussions with the minister’s office” – an opportunity that the PBO apparently didn’t have – and then stated,

This methodology that we used for this analysis is the same methodology, the same source of data, using the [University of Toronto] model that we have used for every other discussion around payroll taxes. So we’ve gone over this with Finance for years, mostly related to the potential for a [Canada Pension Plan] increase. So it’s exactly the same analysis that was used, which is essentially measuring the impact of payroll tax reductions on employment. It’s the same source data that we’ve used, so it wasn’t anything new….We didn’t do a separate analysis specific to this credit [earlier] because essentially it works exactly the same way: it’s a 15 per cent cut in EI premiums and works exactly the same way as any other form of payroll tax increase or decrease would work.

Here is the earlier CFIB analysis that Kelly is apparently referring to, which was published in 2010. The model underlying the analysis assumes that “the Canadian economy [will] close gradually the significant excess economic capacity opened in the recent recession, and…remain at roughly trend growth and ‘full employment’ thereafter” (p. 2).  Kelly’s statement that “we didn’t do a separate analysis specific to this credit” suggests that the CFIB didn’t add any new data or make any changes to the model to reflect what actually happened to the Canadian economy between 2011 and 2014.  So the 25,000-job figure, it seems, was calculated using a potentially faulty theoretical model and out-of-date data.

(It’s also worth noting that this is not the first time that the CFIB has been involved in proposed legislation involving weak data. The CFIB is one of the few organizations that spoke out in favour of Bill C-377, the Conservative government’s blatantly anti-union financial disclosure law.  Bill C-377 was promoted by the federal government as representing the attitudes of a majority of Canadians – a claim based on questionable survey methodology.)

There’s something else worth considering about Oliver’s statements. Before Oliver entered politics, he worked as an investment banker – an occupation which requires considerable expertise in understanding and assessing the accuracy of economic projections.  So the subject which Oliver was testifying about is a subject which he actually should know a lot about. And that makes it even more regrettable that Oliver and the Conservative government are ignoring reliable and well-supported data and instead using questionable numbers in their legislative decisions – numbers that they claim came from “experts”. When even non-experts can see why numbers might be inaccurate, it’s irresponsible for an elected representative to use “I’m not an expert” as a reason to believe those numbers.