Jennifer Victor and Seth Masket recently posted a couple very good posts over at the  Mischeifs of Faction on polarization in Congress. If you aren’t already reading their blog, you should. Both approach the problem from an electoral perspective. There are a few of the regular polarization culprits. For example, districts and states have become more ideologically similar across time. Redistricting has played a role in this as well as migration patterns helped transition the once solid Democratic South into a real two party competition. However, one of their contentions caught my eye: the rules nominating candidates. They argue, in different ways, that state and local party machines gave way to national party organizations in the past 30 years. As a result, political parties have since recruited, nominated, and elected more extreme candidates from both sides of the aisle. This, in turn, led to more polarization in Congress (i.e. high party cohesion and voting records). There is absolutely no question this has happened. Candidates positions, campaign tactics, and even members’ styles all support this contention. Members vote more with their party. Caustic comments and attacks are far more frequent now than prior to the 1980s. No question that today’s candidate is more ideologically extreme than in decades past. There is a lot of research that supports this.
However, polarization is not solely an electoral phenomenon. Evidence of this is in the polarization data itself. As the figure below illustrates there have been two polarized eras in congressional history, the present and the first party era (1880-1910). What makes this interesting is that the two polarized eras span two very different practices in recruiting and nominating candidates. As Masket and Victor point out, the early period was dominated by party bosses and more locally oriented organizations. Candidates were tied closely to their party through patronage at the boss’s disposal. The present period is dominated by national party organizations. In this arrangement, members are more independent (no substantial quid pro quocorruption, though candidates are more likely to ideologically align with their party). In other words, polarization was rampant whether members were independent or not. So there has to be another explanation.

So I’d like to add another culprit in the mix: legislative process. The legislative rules in each chamber are different. So the way that voting polarization is affected by those rules varies. However, there are some regular contributors. In the House, the Rules Committee is the worst instigator of roll call polarization. It has the ability to draft rules that limit minority debate, amendments, and overall influence. It can tailor rules that virtually guarantee major bills will pass, which is really why the House is so much more efficient. Historically speaking, leverage over this committee has had a huge influence on polarization. The below figure illustrates this difference. The shaded areas represent historical moments when party leaders could either directly or indirectly leverage the members on the Rules Committee. There are several ways to influence the committee, but really it boils down to who had committee assignment power. When party leaders controlled it, polarization is much higher. When they don’t, it is much lower.

The Senate is much trickier. Committees were elected by the whole chamber beginning in the early 1900s, there is no Senate counterpart to the House Rules Committee, so structuring debate is mostly done by unanimous consent agreements negotiated by party leaders, etc. However, they can still use these tools to structure debate in the Senate, just not to the extent that it occurs in the House. Similarly, many of the House trends are often carried over into the Senate. Several studies are now showing that polarization and norms in the House are carried over into the Senate. Our own Jordan Ragusa has a very interesting project examining this effect across time and its role in polarization.

There are a couple implications of this. First, the internal-procedural dynamics in Congress play a large role in polarization (as measured by roll-call votes). According to Sean Theriault work, the rules are responsible for nearly 70% of polarization in the House and 60% in the Senate. No matter how independent or partisan a newly elected representative or senator may be, once they enter Congress are forces within the institution, (procedural and financial alike) to wrangle their loyalty. The way that the rules in each chamber influence behavior pushes the parties apart. And second, members aren’t as polarized as we often believe on substantive policy issues. But this point is really splitting hairs. It’s obvious the two parties are far apart. So while there isn’t as big of a gap on final substantive votes normally illustrated by roll-call measures, the gap is still big. It’s like looking at half the Grand Canyon. Smaller than the whole canyon, but still enormous.

This doesn’t suggest that election rules or party organization are somehow irrelevant. I would actually argue they are more responsible for the “real” polarization we see on substantive final votes. But legislative procedures play an important role in the kinds of behavioral norms and voting patterns that emerge particularly those between the parties.