How Can We Explain the Amash Vote to Defund the NSA?

A peculiar thing happened in the House last week.  An amendment offered by Justin Amash (R-MI) narrowly failed on the floor, with a majority Democrats voting for the amendment (offered by a conservative Republican in the Republican controlled House) and a majority of Republicans voting against it.  The controversial amendment would have defunded the NSA program that collects Americans’ phone records en masse.  The program was made public by the now infamous Edward Snowden.  The amendment went down by a narrow 12-point margin.
How can we make sense of the Amash amendment vote?  And what does it tell us about the House of Representatives?  Despite all the commentary about the controversial vote, I’ve seen no rigorous discussion of this in the political science blog-o-sphere.

As an exercise, look at the vote tally and try to find a discernible pattern.  See one?  It’s hard to eyeball, so let’s leverage some basic statistics.  The dependent variable used here is coded “1” if a member voted for Amash’s amendment (to defund the NSA program) and coded “0” if a member voted against the amendment.  I’ll explore five possible explanations: ideology, an ends-against-the-middle pattern, party, terms in the House (as a measure of “outsider” status), and the Patriot Act.  For technical reasons, only representatives who served in the 112th House are included (i.e. no freshmen).

Here are the statistical results.  Positive coefficients indicate an increased probability of voting for the Amash amendment.  Methodological details are at the bottom of the post.


What we see, first, is that ideology is negative and significant, indicating that liberals were more supportive of Amash’s amendment to defund the NSA program.  You read that right.  Liberals were more supportive of a conservative Republican’s amendment to restrict the power of the Obama Administration and, ostensibly, protect Americans’ civil liberties.  Politics something something strange bedfellows.  How can we explain this? Well, liberals have always been skeptical of policies associated with the so-called “War on Terror” (Guantanamo Bay, rendition, the Iraq war, etc.).  Moreover, for conservatives, there’s a cross-cutting issue: the desire to appear strong on national defense.

What about party?  There’s no evidence of a partisan divide in the roll-call pattern.  Amash is not exactly friendly with House Republican leaders, after all.  He famously opposed Boehner’s bid for Speaker.  Moreover, most of the Republican leadership opposed the amendment, with Boehner, Cantor, and McCarthy all voting “no.”

Second, there is clear evidence of an ends-against-the-middle voting pattern in the Amash roll-call.  The pattern was first noted by VoteView blog.  The results indicate that liberals and conservative joined forces in supporting the amendment while moderates—who succeeded in defeating the amendment—opposed it.  For example, on the left, the ACLU urged liberals to support the Amash amendment as did Freedom Works on the right.  Such patterns are rare in a polarized legislature like the House.

However, the results indicate no evidence for VoteView’s contention that this was an “establishment vs. outsider” vote.  Granted, the number of terms served in the House may be a poor way to operationalize “outsider” status.  And freshmen are excluded from this analysis out of necessity.  But junior and senior representative supported the Amash amendment at the same rate once we control for other factors.

The final explanation for the Amash vote is the Patriot Act.  According to the data, 135 representatives who voted on the Amash amendment also voted on the original Patriot Act (which established the NSA program).  By my count, 65% of those who voted for the Patriot Act voted against the Amash amendment while 85% voted against the Patriot Act and fore the amendment.  Even controlling for the factors above, the legacy of the Patriot Act remains statistically significant.  Now I’m not so sure voting on the Amash Amendment was caused by the Patriot Act vote.  Rather, this serves as a key control for unaccounted for characteristics in a lawmakers policy views.

What does this vote tell us about the operation of the House?

The conventional wisdom is that party leaders only allow bills and amendments on the floor when they have the votes to pass the proposal.  The so-called “Hastert Rule” has been the subject of much discussion this year.  So why did Boehner schedule the Amash amendment if it was going to fail?  Well, Boehner has adopted a strategy dubbed “let the House work its will.”  Rather than twist arms—which rarely happens anyway—Republican leaders let the floor decide the amendment’s fate.  But that doesn’t mean that this (failed) vote isn’t in the Republicans interest: quite the contrary.  What party leaders are most concerned with is reelection.  With this vote lawmakers on both sides were allowed to “vote their district” thereby helping their reelection prospects.  Policy outcomes help your party win seats in elections, but so does strategic positioning.

In sum, while the Amash amendment vote was a quirky creature, the vote contains important lessons for how the modern House operates as well as the factors that drive legislative decision-making.

Methods Notes:

The ideology data are available at  Ends-against-the-middle is simply the absolute value of ideology.  Party is coded 1 for Democrats and 0 for Republicans.  The Patriot Vote is coded 1 if a member voted for the original Patriot Act, 0 otherwise.  And terms–a measure of “outsider” status–is coded as the raw number of terms a member has served in the House.