In an earlier post, I argue that The Economic Value of a Law Degree exaggerates the value of a degree to people enrolling in law school now by using a mean rather than a median figure, by excluding taxes, and most of all by failing to account for the cost of acquiring a degree. I conclude that, even accepting all of the authors’ data and methods as sound, the value of a degree to prospective law students would more accurately be characterized as about $100,000 rather than $1,000,000.
In this post I will critique some of the authors’ data, methods, and the interpretations they draw from them.
The authors’ data come from the 1996, 2000, 2004, and 2008 panels in the Census Bureau’s Survey of Income and Program Participation (SIPP).
The sample included 1382 people with law degrees, 58% of whom were actually practicing law (more on this below) which means it surveyed 802 practicing lawyers. Note that none of these lawyers graduated from law school over the past five years, and only very small number have graduated in the last decade (the 1996 and 2000 panels included no such lawyers, and the 2004 and 2008 panels collectively included lawyers from just six graduating classes, of the 52 or so graduating classes included in the sample.)
In addition, the career earnings of the handful of lawyers in the survey who graduated since 2003 are almost wholly a product of extrapolation, based on the career earnings of lawyers who graduated as long ago as the 1950s.
So the first methodological difficulty the authors is that they are using extremely thin or literally non-existent data to extrapolate the career earnings of people who have graduated from law school in the last decade – which is precisely the period during which the cost of attending law school exploded, while the market for (paid) legal services contracted. The authors face this issue head on, by asserting that entry-level hiring in law – which has been admittedly awful in recent years – is cyclical, and that long-term earnings are far more stable.
There are many reasons to doubt this – Richard Susskind and Bill Henderson, among others, have made what seem to me compelling arguments that changes in the legal services sector in recent years are structural rather than cyclical – but one particularly obvious reason to call the authors’ conclusion into question is this: no one seriously contests that, for many years now, the number of graduates being turned out by American law schools has very greatly exceeded the number of jobs for lawyers, taking into account both growth and outflow. In recent years the ratio of new law graduates to available legal jobs has been about two to one.
It’s elementary that, as a labor market becomes more saturated, wages for participants in that market will decline, all other things being equal. The market for lawyers gets worse every year, because every year the surplus labor in that market increases by about 20,000 to 25,000 people. Per current BLS projections, the current decade will produce about 213,000 jobs for lawyers, while ABA law schools will graduate between 400,000 and 450,000 people with law degrees (unaccredited law schools will graduate perhaps 80,000 more).
To ignore this overwhelmingly significant factor when attempting to extrapolate future earnings for people with law degrees, as the authors do, seems willfully blind.
The authors point out that it’s possible to construct narratives in which the economic demand for legal services will increase, which is true. Yet even if one were to accept that those narratives are as plausible as the predictions Susskind, Henderson, and others put forward, I’m not aware of any plausible account of the potential growth in the demand for legal services that posits such increasing demand will actually equal the likely increase in the supply of people with law degrees.
In other words, even according to the most optimistic scenarios, the supply and demand mismatch in the legal services industry is going to continue to get worse for the foreseeable future – the only real question is, how much worse? (The one caveat to all this is that if the federal government put actuarial controls on law school loans, the market for law graduates would in fact correct very quickly, although catastrophically for perhaps half of the nation’s law schools).
This is another way of saying that there’s simply no reason to assume, as the authors do assume, that the fact that a law degree was usually a good investment in 1963 and 1973 and 1983 means that it continues to be just as good an investment in 2013.
And this brings us to the startling fact that 42% of the people with law degrees in the authors’ data set were not practicing law. (The authors excluded people over the age of 65, so voluntary retirement explains very little of this percentage). If nearly half of a group of people who, in the vast majority of cases, acquired law degrees when the market for lawyers was much more robust than it has been in recent years, are nevertheless not lawyers, what percentage of more recent law school classes will actually sustain legal careers? (Again, these latter cohorts are essentially unrepresented in the study).
Now the authors could respond to this by pointing out that they include the extrapolated lifetime income of everybody with law degrees in their study, not just the lawyers. But there are, I think, two major problems with this.
First, this group – people with law degrees who aren’t lawyers – may well be prone to a particularly strong form of self-selection bias. Participation in SIPP is wholly voluntary. People who have washed out of the legal profession would seem to be an especially unwilling group in regard to choosing to participate in such an invasive ongoing survey. The exception, of course, would be people with law degrees who aren’t practicing law, but who have escaped or overcome the stigma associated with having, to borrow Erving Goffman’s typology, a “spoiled legal identity,” by, for example, returning to or finding acceptable non-legal careers.
One critical question raised by the study’s data is, what is happening to the enormous number of law graduates – probably more than half of recent graduating classes – who are not establishing viable careers as lawyers? Is it plausible to think that, on the whole, these people are getting a positive return on their six-figure three-year investment? Law school administrators are desperate to cite any evidence that they are, and I expect Simkovic’s and McIntyre’s Power Point slides will be seen in many law school colloquium rooms over the next few months. Still we need much more data, and in particular much more qualitative evidence, on this matter, given the strong possibility of statistical distortion via self-selection.
Second, and relatedly, the study’s numbers are sobering in terms of the non-pecuniary costs of lawyer over-saturation. Leaving aside future income, it’s important to remember that people who go to law school for non-frivolous reasons generally do so because they very much want to be lawyers.
For example, I know a recent law school graduate who was making around $100,000 per year selling medical equipment before going to law school, but who desperately wanted to be a litigator, even though he realized his attempted career change involved a very big economic gamble. Despite graduating in the top third of this class from a top 50 law school, he spent two years looking unsuccessfully for a real legal job, before economic pressure forced him to go back to a position similar to that which he had before law school.
Now in pecuniary terms, he’s making far more money than the typical college graduate, and indeed more money than he would be making from almost all of the 300 or so legal jobs for which he applied. But, in psychic terms, law school and everything afterwards has been a disaster for him. These are the kinds of significant costs that are being generated by law graduate overproduction which cannot be captured by standard econometric analysis.
In my final post, I will address a few more problems with the paper’s methodology and analysis.