As a baseline, we can start with DJM's calculation that, according to the stats schools reported to the ABA and NALP, 59.8% of the class of 2011 whose employment status was known got full-time jobs requiring bar admission. But it's clear the real number is far lower than that.
Factors that lower it:
(1) Six percent of the class had an unknown employment status. It would be optimistic to assume that a third of those people got real legal jobs.
(2) The default assumptions used by career service offices when they have incomplete data regarding employment status (which is often) is that the missing data is positive. In other words, when there's ambiguity about whether a graduate is working full-time, or is in a long-term or bar required position, that graduate's job will be coded as full-time long-term bar-required. This significantly overstates the number of graduates who have real legal jobs. These assumptions are especially dangerous when applied to jobs with firms of 2-10 attorneys, which accounted for more than two out of every five law firm jobs class of 2011 graduates reported getting. Some unknown percentage of these jobs are law clerk positions, eat what you kill arrangements, or two or three new graduates sharing office space. These distinctions are especially likely not to be made on NALP forms.
(3) Evidence of employment is treated cumulatively. Evidence of unemployment is not. What this means is that a graduate who reports a positive employment status at any point between the spring of the graduate's 3L year and nine months after graduation is treated as employed, period. Graduates who report not being employed will have their status re-checked to see if it has changed.
(4) What counts as non-temporary (long-term) employment by ABA and NALP definitions doesn't actually track with a real-world definition of a non-temp job. A job is considered long-term under these definitions if it has a duration of at least one year. This means judicial clerkships are treated as real legal jobs, even though most of these positions are state district court clerkships, which in palmier days were often reasonable launching pads for legal careers, especially in regard to government jobs, but which under present conditions have morphed for most graduates into one-year way stations on the road to legal unemployment.
Note too that a certain number of "big law" positions in the NALP stats aren't really big law jobs at all. A law professor writes:
How many of the 4,767 jobs (10.7% of all law graduates) that law schools reported graduates of the class of 2011 got with firms of more than 100 attorneys were in this category? Nobody knows. (It says a lot that most current law graduates would consider these "premium" document review positions -- which at least feature one-year contracts, benefits, and offices -- to be "good" jobs).At [mid first tier school], we have quite a number of 2009-2011 grads working with Baker & Hostetler. Those are one-year contract positions, with the possibility of renewal, paying $75,000 per year for document work--but with benefits and a real office. No one knows yet where--if anywhere--those jobs will lead. Yet those are better than average in terms of salary! (And they go into both our BigLaw counts and our salary averages.)And then there are the folks going to the back offices of WilmerHale (Dayton Ohio) and Orrick (Wheeling W Va). They're permanent, indefinite term jobs with starting salaries of $55-60,000 and benefits. But where do those jobs possibly lead? One can't be a discovery lawyer forever--and even discovery is getting more automated.
Also, several hundred "full-time long-term jobs requiring bar admission" were one-year law school funded "jobs" designed to pump up NALP stats at places like Columbia and Virginia.
(5) 2.5% of the class of 2011 reported starting solo practices. This counts as a real legal job if you're shilling for Cooley et. al. Otherwise not so much.
(6) Intentional misreporting on the part of law schools. Confirming whether a school has been lying about the entering credentials of its students is easy, assuming someone bothers to check at some point. Yet a couple of schools have been caught doing just that. Confirming that a school has been intentionally misreporting the employment status of its graduates would be difficult to do, even if the data reported by schools to the ABA and NALP were subjected to some kind of auditing, which they aren't.
All this adds up to the conclusion that, while we don't know what percentage of graduates are getting real legal jobs, it's far less than three out of five. My back of the envelope estimate would be 35% to 40%.