Total Number of Civil Cases and Employment Discrimination Cases Filed in Federal Court: 1972-2006


 So you know, I work on employment discrimination.  Look at this massive plunge in employment discrimination filings.  The left axis is for all civil cases and the right for ED cases (everything classfied as 442 – Tit VII, ADA, ADEA, FMLA, EPA, PDA, 1981, and 1983).  This is the Donohue and Siegelman data in the earlier period (pre 1988) and now a replicaiton and explansion by me, bob, john, and peter all put together in one visually stunning powerpoint slide (kudos, ellen berrey!). 

So, not explained by the previous D&S theory about economic/employment cycles – and it is not all ADA.  its not all 1991 civil rights act damages either.  In fact, the 91 CRA had a backlash effect that is interesting, but for another post. 

We have BOATLOADS of data and can tell you lots about lots about lots.  But, this decline is a bit of a mystery.  Theories?

14 Responses to Total Number of Civil Cases and Employment Discrimination Cases Filed in Federal Court: 1972-2006

  1. jeremy says:

    Can you get more precisely at what type of case has declined, or is it the same across all your classifications?

  2. laurabethnielsen says:

    We collected data from 1988 – 2003 so this new massive decline is not in it. And, oh-so-interesting factoid, no one keeps data on what kind of claim title VII claims are (that is, you can claim race, sex, national origin, or religion in a title VII claim and no one keeps records of which are which). that’s why we went out and did this national random sample of case filings. And now we are analyzing it, but it does not include this most recent drop.

    I will check and see if it is title VII which makes it a mystery, but maybe you are right and it is all one kind of claim (but i don’t think so – i think it is roughly proportional to the types of claims out there).

  3. olderwoman says:

    ED cases rose steeply in the Clinton administrations and declined in the Bush administrations. ED cases initially rose along with all civil cases in the early Reagan administration but then leveled off while other civil cases kept rising. Is there something I’m missing?

    Since it is two series being merged right at about the point of change, you also need to worry about differing definitions in the two series. (I had comparable concerns with merging historical prison statistics with NCRP at 1983, as the two data collection methodologies were different and there was a “take off” after 1983.)

  4. LB–what methodological reason do you have for concluding that a good portion of the decline is not attributable to the ADA, i.e. a decline in state-level disability claims post-Garrett? It seems that may be some of it.

    Also, Title VII religion claims rose in the 1990s because RFRA (the Religious Freedom Restoration ACT) gave employee plaintiffs new causes of action that had been stripped by Employment Division v. Smith in 1990 (holding that neutral laws of general applicability that have the effect of burdening the free exercise of religion are permissible so long as they do not specifically target religious practice).

  5. laurabethnielsen says:

    The 96-04 decline is a decline in all cases roughly proportional to their portion of the case load. Post 04, I can tell you it is the same, except that I cannot tell you WITHIN title VII claims if it is one type or another. That would involve going back out and drawing another national random sample of federal court filings which I have spent the last 3 years and close to $750,000 doing so I am not doing it again anytime soon!

    I can also tell you that religion claims are (get ready folks — all 3 of you reading this — I am about to make news by telling you something that no one knows unless you are on our research team. . . . )

    Religion claims are only 4% of employment discrimination case filings. That’s not 4% of title VII (which includes race, sex, national origin, and religion), that’s ALL employment civil rights cases filed in federal court. And it does not change (differently) over the period.

    Also, Welcome over OW!!!

  6. loosethejuice says:

    How are these cases filed (i.e. does some government agency perform this task? do private individuals hire lawyers and sue)? I don’t really know anything about this line of research at all, so I have no clue. The business I am in, however, is thinking about the influence of political processes on policy outcomes. And, if I didn’t know any better, I would think that your line chart here qualifies as a policy outcome, with some likely political process driving it.

    Conditional on the answer being an agency (or, without a huge stretch, we can think of the courts as an agent as well), there may be sufficient conditions to cause a decline purely due to political reasons, even in the face of the typical macroeconomic effects. Further, it may be the case that political control may not be explicit, yet still have a direct effect. One of the more profound conclusions of the body of agency discretion literature is that, from the prospective of agency discretion, a whole lot of legislation and no legislation at all are observationally equivalent (for example see Weingast and Moran “Bureaucratic discretion or Congressional control.” [i]Journal of Political Economy[/i]). In one case, the agency is being reigned in by legislation. In the other, the agency is doing exactly what politicians want it to. It may be the case that not all politicians want the actual policy changes, but that the actions of the agency are being allowed by some other gatekeeping actor (such as congressional subcommittees), or are responding to particular actors (such as veto actors) which may or may not be changing in ideology at the same time as other parts of the government.

    This literature would likely say something along the lines of “the reason there are fewer ED cases filed now as compared to Clinton, is because politicians wanted fewer cases filed today, as compared to Clinton.” Depending on which piece you read, the title of the named politicians varies, but is likely to be enlightening to think about who would control that process and what their incentives are. I bet if you put in a variable for presidential ideology, in any case, (such as Poole and Rosenthal’s common space scores it would have an effect. An ocular trauma test indicates that is likely to be significant.

    The work of my advisor (and now, by proxy, my work I guess, lol), takes agency outputs and models them as “regimes.” The idea being, for a given regime a particular actor is more likely to have an effect on output, due to the legislative subgame that is being played in Congress. Let’s assume that the Judicial subcommittee in the House and Senate has jurisdiction over most of the legislation regarding this policy area. Given this, my data shows that just pre-Clinton, the relevant actor was the moderate House and Senate floor medians. At the onset of the Clinton era, for the first couple of years, the agency in question is most likely responding to a quite liberal Judicial subcommittee (-.33 on the Common Space Scale. Bush, for reference is .538 on a -1 to 1 scale). Following this liberal jump, we again had moderates from the floor directing agency actions. Again following this, the regime type switched back and forth between responding to somewhat liberal subcommittees and the ideology of the President. This would likely account for the large spike in Clinton (which makes sense) and the mild decline under Bush (which doesn’t make sense).

    Fuck, I should be working on something else.

  7. laurabethnielsen says:

    they are talking about us over here: and I posted lots of the answers to loose the juice –

    short answers — regulatory agency is the EEOC. Individuals must obtain a right to sure letter (with some exceptions) before heading into federal court.

  8. laurabethnielsen says:

    I don’t think this has to do with an underlying change in the volume of discrimination. remember, since these cases have first gone through the EEOC, there is a lag. Also, the president’s influence in this is appointing the head of the EEOC who primarily signals what kinds of cases they will take as an agency. Clinton wanted to get some good ADA cases through the courts to try to make some good precedent. But the EEOC takes a very, very, very small number of cases. Mostly they process your form and send you to federal court. Well, they tell you that you MAY go to federal court. Any guesses as to how many people who CAN go into court actually DO?

  9. onemorecup says:

    Pursuant to the ‘rules of writing’ one should spell out for their audience the names that the acronyms represent rather than leave one guessing. Just something worth mentioning.

  10. laurabethnielsen says:

    Tit VII = Title VII of the Civil Rights Act of 1964
    FMLA = the Family and Medical Leave Act
    PDA = the Pregnancy Discrimination Act
    ADA = the Americans with Disabilities Act
    91 CRA = the 1991 Civil Rights Act
    EPA = the Equal Pay Act
    ADEA = the Age Discrimination in Employment Act
    D&S = Donohue and Seigleman – the people whose study we are replicating and expanding

  11. laurabethnielsen says:

    You can now read the entire report here:

    Click to access nielsen_abf_edl_report_08_final.pdf

  12. free seo software…

    […]Total Number of Civil Cases and Employment Discrimination Cases Filed in Federal Court: 1972-2006 « Controlling Authority[…]…

  13. grams to ounces…

    […]Total Number of Civil Cases and Employment Discrimination Cases Filed in Federal Court: 1972-2006 « Controlling Authority[…]…

  14. ipad says:


    […]Total Number of Civil Cases and Employment Discrimination Cases Filed in Federal Court: 1972-2006 « Controlling Authority[…]…

Leave a Reply

Please log in using one of these methods to post your comment: Logo

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

Google photo

You are commenting using your Google 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 )

Connecting to %s

%d bloggers like this: