How do we measure job vacancies – beyond Kijiji

The Federal Government and Minister Kenney got a pretty good ribbing over their flawed methodology in measuring job vacancies – and with good reason. Anyone who has used Kijiji for anything knows how unreliable it can be. The difference between Statistics Canada’s method and one that includes Kijiji is pretty easy to communicate.

But there is another method from the CFIB that gets some attention, and different results than Statistics Canada. I usually talk about the number of unemployed workers for each job vacancy, but the proportion of job vacancies to total labour demand (job vacancies + jobs) is a very useful measure too. CFIB finds a job vacancy rate of 2.5% for 2013 Q4 and Statistics Canada reports 1.3% in December 2013.

This is sometimes presented as a puzzle, but the different result is mostly because the CFIB measures something different than Statistics Canada. The CFIB is very transparent about what exactly they are measuring, and why they think it’s important. This measure adds information to the policy debate, and that’s great. I think a better understanding of the differences enhances the debate.

First, the CFIB only surveys private businesses. That’s useful for their membership, and understanding unmet labour demand in the private sector. Statistics Canada includes public sector jobs. The CFIB suggests that job vacancies are lower in the public sector, but we don’t have conclusive data on that. Statistics Canada tells us that the job vacancy rate in the field of Education Services (which is 90% public sector) is very low – only 0.4%. But Public Administration (100% public sector) is at the national average of 1.4%. So it’s unlikely that this could explain the whole difference between the two numbers.

But wait, it’s not possible that low public sector job vacancy rates would explain the difference in results, because the CFIB method finds more job vacancies in the private sector (295,700) than Statistics Canada finds in both the public and private sector combined (202,500).

So what explains the difference? Well, Statistics Canada requires that an actual position exist for it to be considered a job vacancy. CFIB does not require that a position exist, or that a business be actively advertising to fill that position. All that must exist is a need. This is clearly explained in the CFIB’s methodology section of their regular job vacancy releases. This is a broader definition of labour demand, that tells us something useful. It is comparable to including discouraged job seekers, who aren’t actively looking for work, but express a desire to work if a passive search connected them to a good match.

So Statistics Canada’s job vacancy measure is to the CFIB job vacancy measure as unemployment rates are to underemployment rates. For the most part. Maybe when we get the upgraded job vacancy survey that Minister Kenney has promised to fund, we’ll also get a public / private breakdown of statistics so that the two measures can be better compared.