Thursday, 9 December 2010

Proxies for Happiness

There’s been a lot of talk recently from the government about measuring happiness to use it as a metric of success. Now perhaps the ultimate aim of all governments ought to be to increase happiness, but given that it’s such a hard thing to measure objectively, most settle for some economic measure (GDP or GDP per capita) as a proxy.

This raises many questions about whether GDP is a good proxy for happiness (for one it ignores income disparity), and whether some alternative measure would be better suited, say by conducting some kind of happiness survey. Rather than answering this thorny question, I want to take a look at a somewhat simpler problem - measuring career satisfaction for new graduates.

Earlier this year whilst doing some research for a potential new startup, I spent some time looking at university careers services. These are the departments within universities that are responsible for helping students find jobs once they graduate. These departments all seem to be driven by one key statistic - the number of graduates in employment 6 months after graduation (of those looking for employment, so excluding those pursuing postgraduate study etc). This is often published in prospectuses and university websites as a means of attracting students, and presumably is also used internally to measure the success of the careers departments. The use of this statistic is also reinforced by the Higher Education Statistics Authority (HESA) which conducts an annual survey on the destination of graduates 6 months after graduation.

Now this certainly isn’t a bad statistic to measure, and combined with the salary distribution information most careers departments also publish, it should give any potential students a fair idea of what to expect from a given university. What these statistics completely fail to measure is whether these students have ended up in jobs they are happy with. You could argue that in the current economic climate, students might well be happy to have any job, but in the longer term its probably better for both employers and employees if they’re not tied together solely by economic necessity.

So this takes us right back to the problem of measuring happiness. Except that in the case of jobs there’s a very simple proxy for happiness. Namely, the length of time a person stays with their employer. Or to make it a bit easier to digest, the percentage of people that remain in their job 2 years after graduation (or whatever time interval you choose). There’s an assumption here that the majority of people who leave their jobs do so because they’re unhappy (for whatever reason), which seems reasonable. Maybe they’re unhappy with pay, progression, location, flexibility, hours etc - but clearly something makes them move on. It’s possible that they might not have left of their own volition, but again it seems reasonable to assume that redundant or fired employees aren’t filled with joy at their situation. The time interval needs to be reasonably short since we’re effectively also assuming that happiness when they leave is representative of the whole period of employment, which won’t be true for longer periods (e.g. somebody who leaves after 5 years may well have been very happy for 3 years).

There’s obviously a cost associated with universities tracking this kind of data as it would require them to contact all their alumni two years after graduation (though doing this online would reduce costs). However, given the increasing squeeze on funds at universities, might it not be prudent to stay in regular touch with alumni as possible future benefactors? Using a hapiness metric would incentivise career services to place students in jobs that they’re likely to be happy in, and it wouldn’t surprise me if they find happy alumni are more inclined to reach for their chequebooks when the alma mater comes a-calling.