Below is a chart I made during lunch today that plots the number of public laws passed by October 1 of the first session of Congress vs. the number of laws passed during the remaining 15 months of the Congress. It plots the past 19 Congresses (93rd-111th) and a predicted value for the 112th (labeled in purple). Unified government is labeled in green and divided government is labeled in red (the 107th Congress, when Jeffords switched parties, is labeled as divided). The lot does not differentiate between major and trivial legislatio n; it is strictly limited to the number of public laws.
1) In general, only a small percentage of legislation is passed prior to October 1 of the first session. Over the 20 Congresses in the plot, an average of only 16% of the total legislation in a Congress is passed prior to 10/1 of the first session. This is both surprising and not surprising. It’s not surprising because it takes time to develop legislation and move it through the chambers; committees need to hold hearings, write bills, mark them up, and then jockey for floor time. On the other hand, 16% is still a lot lower than I would have guessed, especially since the second session almost always wraps up its second session work prior to the election in November.
2) There is an unsurprising positive correlation between the two variables. Simply put, Congresses that pass more legislation prior to 10/1 first session tend to pass more legislation post-10/1. Not rocket science. And also not that important, because a huge part of the absolute volume difference is institutional: we haven’t filtered out trivial commemorative legislation or other non-controversial items, and my guess is that they drive a lot of the result. Prior to the 1995 ban on such legislation, commemoratives were a sizable chunk of all public laws. So a high volume of pre-10/1 legislation is likely to correspond to a high volume of post-10/1 legislation — there were just lots of non-controversial public laws at all times. And that’s the case: all of the highest volume years are pre-1995; all but one of the lowest are post-1995.
3) The 35 public laws passed by the 112th Congress is not a huge X-axis outlier. It’s the second lowest absolute number, but the pre-10/1 output of the 112th is not all that much less than other post-1995 divided Congresses. As mentioned above, comparing the 112th to a pre-1995 Congress raises the issue or the commemorative legislation. I haven’t gone back and filtered that out of the older Congresses, but I suspect it would have a large effect.
4) The outliers (may) tell an odd story. Notice the five biggest outliers from the trendline, circled in red. Those are instances in which either a proportionately large amount of legislation was passed after 10/1 (96th and 106th) or a proportionately small amount of legislation was passed after 10/1 (103rd, 110th, 111th). What’s odd about those Congresses is that they all preceded changes from either unified to divided government (96th, 103rd, 111th) or from divided government to unified government (106th and 110th — although the Jeffords switch quickly reversed the 106th point). In fact, in the entire sample. on the 94th Congress preceded a change from unified/divided government and did not stray from the trendline.
I don’t know what to make of this. I guess we could make up a story about the coming election tides either locking up Congress or getting it into gear. But the pattern doesn’t match up how we might expect: if that story was true, then we’d expect a unified government to get into gear when they were about to lose control, and a divided government to lock up when one side or the other looked poised for victory. Instead, we get lockup in both the 110th and 111th, one of which created unified government and the other of which created divided government.
So perhaps it’s a spurious finding. But it’s a darn interesting one. If anyone has seen any research on legislative productivity as it relates to perceived coming changes in unified/divided government, let me know.
