13 February 2016

An update to an older problem

A problem I identified among, mostly liberals I guess, I believe has a more firmly identified cause than it did some years ago when I first wrote about it. 

The problem was voter turnout and associated beliefs about "uninformed" voters screwing things up. I categorize most of these as "misinformed" voters, because most of these are voters who believe they possess a good deal of information about complicated subjects and tend to score well on tests of basic political knowledge. Uninformed voters, people possessing very limited levels of political knowledge, tend not to vote. Because they are not interested in politics usually. This is a problem for the theory that voter turnout produces a skewed outcome of some sort, in that the biases of "informed" voters already exist and are generally reflected in the vote.

A clearer statement of this appears to work more as follows: "uninformed voters" are "Republican voters", and that supposedly there is a legion of "informed voters" who don't vote (which is generally not the case), who would vote for Democrats. I'm dubious of this assumption being correct either. The available research on voting participation in the US in particular suggests that the vast majority of people who vote, particularly in primary elections as now, but also in general elections are reliably more partisan than people who do not vote. This is less about information, though partisanship can skew information considerably, and more about political parties having stronger incentives to turn out their voter bases than to try to convince moderate voters to vote for them, voters who are also reliably less informed and thereby somewhat difficult to appeal to, or to get weak partisan voters of your opponents to defect to your side or at least not show up at all, voters who will require somewhat more convincing still. In other words, it's becoming a race to appeal to strong partisans to argue among themselves about what dastardly plan they won't be able to accomplish because those evil people over there are preventing it and dammit we need more of us, and then point fingers at how terrible those people are on the other side. Or even politicians who aren't "pure" enough (as with the Sanders-Clinton infighting that occurs online).

The effect of this however on voter turnout isn't "more people will vote Democrat if only they would show up", because there's less evidence for that prognosis, particularly as the political parties become more extreme. It seems to be instead fewer people give a damn in the first place because they feel disconnected from the more partisan and more extreme voter bases of political parties. General elections like the Presidential election cycle will probably still turnout "moderate" or even "establishment" voters who are less partisan than ordinary voters. Most other forms of elections may become increasingly extreme and polarized, and eventually have fewer voters (not yet, more work would be required to be more partisan than now). This theory seems to explain some of the chaos that has emerged so far as a narrative among the political primaries being so divorced from "establishment" demands, in that the partisan activists are very displeased with the performance of their chosen political parties and believe they would do better if they were only more extreme and purer than they currently are. In some way. There is little or no historical evidence for this theory of course. In fact extreme partisans in large turnout elections have a very strong pull historically toward "this will cost us the election" or "this could splinter and fracture our political party" or "some third party or independent candidate will get a bunch of votes". None of which seems likely to be optimal to the extremists but still they persist in their motions.

As a minor member of a disloyal tribe of partisan voters (small l libertarians, or left-leaning ones), this is all quite amusing watching the major political parties trying to eat themselves alive by arguing their way into factionalism and extremism. And I gladly cheer their efforts on if it means more factions emerge eventually and perhaps more choice becomes available to capture factions of voters unmoved by those on these current extremes (as I am not). But how this impacts voter turnout in some positive manner seems far from clearly advantageous for either side. For the following reason: right now the public still seems on balance to be roughly split in so far as that it tends toward more socially liberal policies (they want less interaction between values and laws), but also tends toward more economic conservatism (they want government to do at least somewhat less than it presently does in the economy overall, if not specifically). At present neither political party represents this demand sufficiently at a national level (one could argue Democrats during the period from 1930 to 1970 mostly had control this state of affairs as it was a more socially conservative party then nationally, and perhaps Republicans in the 1980s and 90s as the national mood was more socially conservative). This dynamic will also shift back and forth on both counts, or at least it traditionally has. Neither party likely possesses some sort of natural monopoly on voters that can persist for long periods of time. This suggests that polarized elections of lower turnout will eventually produce unstable swings between parties as little of note may get done, or that what gets done is immediately demanded to be undone. Non-polarized elections with higher voter turnout will tend to produce more stable swings as some number of people in the middle will be able to sway elections if they feel one party or the other has too much or too little influence.

There can be an argument that voter suppression efforts to depress turnout have a general partisan bias. But even this is not guaranteed. Evidence suggests that the attempts to use voter ID laws in Pennsylvania for instance may have harmed Republican vote turnout more than Democratic vote turnout in the state (or, more likely that it had little partisan effect versus the fact that it was not expected to be a competitive state). It is also not abundantly clear that all such laws have the effect of swinging partisan districts between parties, or dramatically reducing partisan turnout for an opponent in a general election. Voter restrictions based upon felony convictions are a possibility to be of a partisan effect. Voter restrictions that end up with a disproportionate racial bias also. The evidence also suggests that the strictest voter restrictions of these types (for example voter ID laws) end up with perhaps a 2% reduction in voter turnout. This results in perhaps an additional 1% of Democratic voters by margin on average, which is unlikely to impact most races (it will impact some surely, but will also push some number in the other direction). Partisanship effects for something like primary elections have had far more dramatic impacts upon voter turnout however, as voter turnout is considerably lower than a 2% reduction.

To be sure, I tend to oppose most restrictions on voting. For example, I am in favor of allowing ex-convicts to vote in any state that currently restricts it (Maryland just changed its laws on this point, overriding a Governor's veto). I tend to think strict voter ID laws are generally not very useful. Very few people will attempt to vote fraudulently without them. Much less than this 1% of partisan voter suppression being achieved with them. Which suggests these laws are unjustified or unnecessary for electoral fairness. Nevertheless the idea that such laws have significant impact on turnout and the results of elections (particularly Congressional elections) seems spurious versus other causes, partisanship for example, or favorable redistricting efforts in partisan directions, or incumbency biases, and so on down the line. Few elections turn upon a 1-2% turnout gap. The average margin in House races tends to be over 30%. Senate races are about 20%. Approximately 30 House seats and 3-5 Senate seats will be under even 5% each cycle. Maybe 60 House sears and 8 Senate races are decided by fewer than 10 points. This is not significantly different in off year cycles like the midterms. This should not be encouraging that low voter turnout is the problem for some vast quantity of annoying Republicans getting elected if one is a liberal Democrat.
Post a Comment