New evidence that lobbying affects legislative outcomes

LOBBYING

Lobbying influences the agenda but does not buy votes

As state legislatures across the country wrap their annual sessions, commentators are trying to explain why some bills passed while others failed. A common element in these explanations is lobbying — namely, that the advocacy efforts of organized interests for or against a bill led to its enactment or defeat.

A key challenge in studying the link between lobbying and outcomes is data scarcity. In most American legislatures, organized interests are not required to publicly report on which bills they lobby and the positions they take on those bills. Consequently, we rarely know which bills experience lobbying by which interests and the preferences of those interests lobbying.

Specifically, we draw on lobbying reports from three states — Colorado, Nebraska, and Wisconsin — where lobbyists are required to report on which bills they lobby and whether they are lobbying for or against each bill. Together with the legislative histories for each of the over 26,000 bills in our data, we can assess whether bills that experience more favorable or unfavorable lobbying are more likely to be enacted or fail to become law, respectively.

Our results indicate that lobbying is related to bill outcomes.

Pooling our data from all three states, we find that bills experiencing only favorable lobbying are 11 percentage points more likely to be enacted relative to bills experiencing no lobbying. Conversely, we find that bills experiencing only unfavorable lobbying are 26 percentage points less likely to be enacted than are bills experiencing no lobbying. Finally, when bills experience both favorable and unfavorable lobbying, we find that they are 13 percentage points less likely to be enacted as compared to bills experiencing no lobbying. These results suggest not only that favorable and unfavorable lobbying affect bill outcomes, but also that the effect of unfavorable lobbying is stronger than that of favorable lobbying — a finding consistent with a status quo bias, or the notion that policies are easier to maintain in their current form than to change.

We also use our data to test two popular explanations for how lobbying is thought to influence outcomes. First, we assess the role of agenda setting, or organized interests’ power to influence which bills advance further on the legislative agenda. To test the agenda-setting explanation, we examine whether favorable and unfavorable lobbying are associated with bills’ likelihood of reaching intermediate steps of the legislative process, such as emerging from committee and receiving consideration by the full chamber. Our results are similar to those in our main analysis where we focused on bill enactment; bills experiencing only favorable lobbying are more likely to advance through important steps of the legislative process, while those experiencing only unfavorable lobbying or both favorable and unfavorable lobbying are less likely to progress.

Second, we evaluate the role of vote buying, or organized interests’ ability to convince legislators to change their votes to make them consistent with the interests’ own preferences. To test this explanation, we assess whether legislators who cast votes on bills at the committee stage change their votes once the bill reaches the floor in ways consistent with new lobbying that occurs in the interim.

For instance, if the vote buying explanation underlies the effectiveness of lobbying, we would expect legislators who voted for a bill in committee to vote against it on the floor if new interests began lobbying against the bill between the committee and floor votes. Unlike our agenda setting analysis, we find no support for the vote buying explanation, as new lobbying between a bill’s committee and floor votes does not induce legislators to change their votes.

As with all studies, our analyses have limitations. For instance, while our novel data provides rarely observable information on lobbying activity on specific bills, we are unable to make causal claims about the effect of lobbying on bill outcomes. We try to minimize this concern with survey data we collected from lobbyists in the same three states that discount alternative explanations for our results, such as that lobbyists lobby for or against bills they expect to pass or fail, respectively, rather than affecting the outcomes themselves.

Additionally, because we cannot analyze similar data in other states and at the federal level, we are unsure to what extent our results generalize to Congress and other state legislatures. However, given the rich variation in the demographic and institutional characteristics across Colorado, Nebraska, and Wisconsin, we are optimistic that our findings indicate a generalizable relationship between lobbying and legislative outcomes as opposed to patterns unique to these states.

Related posts:

When I was implementing the Genetic Algorithm (GA) and Genetic Programming (GP) library, Jenetics, it was clear that I needed a way to change the PRNG implementation the GA was using during the…

Fragmentation occurs when an IP datagram traverses a network which has a maximum transmission unit (MTU) that is smaller than the size of the datagram. Take for example a standard Ethernet datagram…

Doing this project presented a few challenges. One of the main things that could have been done better is the filming. I did not notice until I was editing the video, but the audio levels were a bit…