Legislative leaders skip way more votes than non-leaders.
Let’s continue this study of absenteeism in the Utah legislature. Earlier, I showed that Utah legislators skip lots of votes. Then, I showed that legislators miss some types of votes more than others. Next, I gave a list of which legislators have the worst attendance records. Now, let’s wrap this up by comparing legislators who miss lots of votes to legislators who miss few votes. Are there predictable patterns?
There’s an easy way and a hard way to study this. The easy way is easy to understand, even if you don’t know statistics, but it might give inaccurate results. The hard way is harder to understand, but it gives better results. Let’s look at the easy way first.
The easy way (but it might be wrong)
I’ve calculated how often each legislator missed votes. (Raw data are here.) These are percents: During the 2011 session, what percent of the time was each legislator absent instead of voting yes or no? In this analysis, I’m looking only at absenteeism in the Utah House.
In the table below, I compare average absentee rates for various groups. For example, there are 8 leaders in the Utah House. Four are Republicans and four are Democrats. Overall, leaders miss 12.6% of their votes, but nonleaders miss only 6.4% of their votes. Here’s the complete table:
|Group 1||Group 2||Difference|
|Low floor power||5.7%||High floor power||7.7%||+2.0%|
|Sponsor few bills||6.7%||Sponsor many bills||9.0%||+2.3%|
In each row of the table, the difference is boldfaced if it is outside the margin of error and italicized if it is not.1 But remember, this “easy” analysis has some flaws that I’ll address in a moment.
The difference between Republicans and Democrats is interesting, but it’s not large enough to be meaningful. The same is true of any other difference in italics.
Seniority seems to play a role. Rookies are those who have served 0, 1, or 2 previous sessions, placing them in the lowest quartile. Old timers have served at least 8 previous sessions, placing them in the highest quartile. Old timers have a much higher absentee rate than rookies.
You might think you would miss more votes if you were sponsoring many bills. After all, you may need to leave the chamber to push your bill through the other one, or to meet with people about your bill, and so on. It turns out the difference between those who sponsor lots of bills (8 or more, the upper quartile) and few bills (3 or fewer, the bottom quartile) is within the margin of error, though. (I’m not counting floor sponsored or co-sponsored bills here.)
So, this table shows that leaders and old timers miss more votes. It may also be that Republicans and legislators who sponsor many bills also miss more votes, but those estimates are within the margin of error.
Okay. That’s the easy way to do the analysis. Here’s the problem with it: Many of those variables move together. Do old timers miss more votes because they are old timers, or is it that old timers are also more likely to be (a) leaders and (b) the sponsors of many, many bils? The effect of being an old timer might actually be driven by something else. So let’s move to the “hard” analysis.
The hard way (but it’s more likely to be right)
The “hard” way to do this analysis involves some statistical techniques that I won’t explain here. These techniques can help us isolate the effect of each of the variables listed above, other things being equal. That “other things being equal” bit is key. We can estimate the effect of gaining an additional year of experience, for example, while holding constant other things like your leadership status, how many bills you’re sponsoring, your party, and so on.
When I use these techniques, I find that the effect of seniority mostly goes away. It appears to get swallowed up by the effect of being a leader, which has a massive effect.
I also find that the number of bills you sponsor continues to have no effect on your absentee rate. Strangely, however, I find that if you increase the number of bills you “floor sponsor,” your absentee rate goes down. (It’s called “floor sponsoring” if a Senator approaches a Representative and asks him to sponsor his bill through the House. The Senator is the main sponsor, and he already guided the bill through the Senate; the Representative is the floor sponsor, and his only job is to get it through the House, with the main sponsor’s help.)
It’s a bit counterintuitive that additional floor sponsorship would cause you to miss fewer votes. I would expect the opposite. The more bills you’re shepherding, the more you might need to leave the floor to build your coalition. I’m still puzzling through this.
I’m still collecting some additional variables that I haven’t mentioned here, and if I have any additional findings that seem interesting I’ll post them later. In the meantime, the clearest finding is this: Legislative leaders skip way more votes than non-leaders.
Update: See “Which legislators run the most bills?” for further discussion of how bill sponsorship activity might hurt attendance.