Explaining high drop-out rates

College drop-out rates are soaring in America and most people seem to blame high tuitions for that. In this post I will compare the U.S. with other countries to see whether there is indeed a connection between tuition and college drop-out rates.

The post was inspired by a conversation I had with a friend of mine about higher education. When trying to explain markedly different drop-out rates in our respective countries, the topic of tuitions came up.

The data is for 2011 and is obtained from the OECD. I have four datasets for the countries:

  • teritary education completion rates,
  • teritary type-A education completion rates: type-A education refers to largely theory based, research-focused programs and/or programs with high skill requirements (like medicine),
  • annual average tuition fees at public teritary type-A institutions,
  • annual average tuition fees at private¬†teritary type-A institutions.

While the tuition data is for teritary type-A institutions, I have no doubt that most countries have roughly equal tuition fees for “normal” teritary institutions and/or that most teritary type-A institutions are also “normal” teritary institutions.

Doing all combinations of regressions with the above data that make sense will not yield any result that would support the hypothesis that tuition fees have an effect on drop-out rates. The coefficient of tuition fees (whether public or private schools are used) is not statistically significant from 0 in any of the regressions. Using logs of the data won’t change that.

I would have considered more complicated functional relationships, specifically I thought a quadratic relationship might make sense, but when I plot out the data it just doesn’t seem right. And a quick quadratic regression confirms that.Tuition vs. drop-out rates

The point is however I might try to torture that data, I cannot get a significant relationship between tuition fees and drop-out rates. (A grouped logit or probit model doesn’t show a significant effect either.)

Neither the Park, Glejser, White and Koenker-Bassett tests nor visual inspection indicates that there is heteroskedasticity, therefore I did not experiment with transforming the data accordingly.

The conclusion seems inevitable: at an international level, tuition fee differences cannot explain drop-out rate differences. Different drop-out rates across countries are therefore the result of other underlying factors. One reason could be different admissions requirements: certain countries are quite egalitarian and prefer to give most students the chance to attend college, yet many drop out in the first one or two semesters because they realize college is not for them. Other countries have much more selective admissions.

Another reason may be that some countries (and this definitely applies to the U.S.) tend to make college education out to be the only possible way for children to succeed in life. We leave no other options to our kids. Therefore, a lot of kids who are not college material (i.e. they’re not hard-working enough, they do not have enough self control or discipline, they are not prepared/intelligent enough, they’re not motivated) are pushed into college just so they can drop out.

This doesn’t appear to be a revolutionary find, others have written about it. Yet whenever I dare bring it up, I seem to get negative responses. I still stand by my conclusion: U.S. drop-out rates are high because we send way too many non-college material kids to college.

On the other hand, the question comes up: is this good or bad? Of course, there is a lot of inefficiency because all the administrative and teaching resources those future drop-outs consume are essentially wasted. But then, giving everyone the chance to at least try college can help us find “diamonds in the rough”. It can help us discover students who did not quite succeed in school before but who can soar in college, since college is in many respects much different from high school (Roland G. Fryer anyone?). These people are rare of course, but perhaps it is worth some inefficiency and higher drop-out rates to give them this opportunity. For if we were to tighten admissions standards, they could probably never attend college. Also, inequality and social mobility are already terrible in the U.S., thus we do not need to exacerbate these further.

Bottom line: while strengthening vocational education and apprenticeships could do wonders in the U.S., we should perhaps stop worrying about high drop-out rates.

Funny fact to end this post with: socialist Sweden with completely free teritary education has similar drop-out rates as the U.S.

Source: OECD for all data (1, 2).

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s