Thank you.
I'm going to look at something that's much more specific. I'm interested in seeing if I can explain some of the variation in completion rates across provinces and trades over a period of around about 15 years.
The problem is motivated by a fair bit of media attention over the last 10 years about a perception that there's a shortage of skilled trades, often in construction.
Andrew mentioned quite a bit of this. Registrations are up. My numbers are a little bit more out of date than Andrew's, but the number of new registrations in apprenticeship programs went from 30,000 in the early 1990s to 80,000 in 2007. If you then look at completions, you just don't see increases of the same order of magnitude. So the obvious implication is that the completion rate has declined over that period. I have a number here from a paper of Andrew's in 2005 that talks about a fall from 64% in 1982 to 39% in 2002, although there has been some increase over the last few years, as Andrew mentioned.
When you look at the data at a less aggregate level, you see quite a bit of variation across trades and provinces. In the handout, I gave you some numbers for bricklayers in 2002: in Alberta the completion rate is around 50%; in B.C. the completion rate is around 14%. The number I'm looking at there is not completions over total registrations, but completions over new registrations, say four years previously when the program would have begun.
If you look at the requirements for apprenticeship programs across provinces and across trades, you also see variation. Again looking at bricklayers in 1999, to start an apprenticeship as a bricklayer in Alberta you would have needed a grade 9 education and then you would have been required to do about 5,500 hours of work experience; to enter the same program in B.C. would require a grade 12 education and then about 5,000 hours of work experience. There are similar variations in the amount of training you have to do and the format of that training across provinces for the same trades, and obviously, of course, there are differences in competencies across trades.
What I'm aiming to do is exploit that variation in the apprenticeship programs across provinces, to see if that's related to the variation in completion rates across provinces and to see if that will tell us something about an optimal way to design these apprenticeship programs in order to raise completion rates.
The ways in which these programs differ is whether or not certification is mandatory to work in a trade; the length of the work experience term, as I mentioned; the amount of formal training; the method of delivery of that training; and the level of education required to enter the program. So those are the features of the program I'm looking at.
I also look at the population of apprentices in these trades and provinces: the age structure of apprentices in each trade and province, and also the sex composition as well. I am asking whether those are related to completion rates.
Finally, it is often argued that employment instability is an important factor in apprenticeship completion rates, so I also look at unemployment rates across provinces and by aggregated trade indicators.
The data comes from a bunch of different sources, and I won't bore you with where the data is from. It is in the paper, if you are interested.
As a summary of results, the presence of mandatory certification is associated with about a 10% higher completion rate than in trades without mandatory certification. This is probably not surprising. If you need to complete the program to work in a trade, the incentive to complete the program is obviously much greater, and we would expect to see apprentices complete more often.
Does that mean that mandatory certification is the right policy? It may not be the case. If the issue is that there is a shortage of people working in this trade and non-certified workers are reasonable substitutes for certified workers, then mandatory certification may not help alleviate a trade shortage.
Looking at the length of programs, there is really no evidence that the length of program is related to completion rates at all. It's not the case that longer programs have lower completion rates, as you might expect. There are a number of reasons why that might be the case. Apprentices who choose those programs may be different from apprentices who do not choose those programs. So again, we're not randomly assigning length of programs to individuals. It may be that those who choose these programs understand fully that these are long programs.
The one thing my results don't speak to is whether this is a barrier to entry. It could be that long programs act as a barrier to entry. If you know it's going to take you five years to qualify as an electrician, you may never enter the program in the first place. So my results don't speak to that. It could be that shortening programs may have a positive effect. But again, the trade-off is that if you shorten the program, there's less training for the individual, and then whether or not they gain the competencies that the labour market requires could be an issue.
It's often argued that having technical training delivered by block release is a barrier to completion. I find no evidence of that at all. The provinces and trades with training delivered by block release have no lower completion rates than others.
There's a negative effect of average age on completion rates. Apprenticeship programs with younger apprentices typically have higher completion rates. We might expect that younger apprentices are less likely to have family commitments that are going to cause them to drop out of programs and so forth. This effect is fairly small, and similarly, a positive effect of the “percentage female” on completion rates again is a fairly modest size.
The one thing that does show up as important is unemployment rates. Trades for which there's a high unemployment rate over the period for which the apprentice will be taking the program typically have lower completion rates. We would expect that if you are having difficulty keeping a job, then you're not able to accumulate the hours of work experience you need to complete, so the completion rate is lower when there is a higher unemployment rate in that trade and province.
Thank you. I hope I didn't go over too much.