Regarding Democratic Party Approval Hitting Record Low

The Washington Post's Jose A. DelReal writes:

The Democratic Party's approval numbers have hit their lowest point in at least two decades following an electoral drubbing that gave Republicans control of both houses of Congress, according to a new poll released Wednesday.

The Gallup survey has the Democratic Party pulling a 36 percent favorability rating, a 6-point drop since September.

The same poll shows the Republicans tying a 2-year high in approval at 42%. Further, by a 17-point margin, the public prefers Congress to steer the agenda, not the President.

Obviously, Republicans will be pleased with this poll. However, there are some things to keep in mind.

1. I have long held that, in the wake of a 'big' event, slight changes in response rates for each side of the aisle can have a (temporary) distorting effect on polls. Especially with response rates getting low, just a 1% increase in the chance that a person called will respond due to them being psyched about things can change the measurement tremendously, as can a decrease in the willingness of those turned off by the recent turn of events. This is one of the reasons why there are frequently 'bounces' in approval ratings or horserace measurements after events such as convention acceptance speeches or State of the Union addresses. Such distortions do not reflect a change in the underlying attitudes, and relatively quickly fade. Odds are, part of the results in this survey are due to such effects.
2. Approval rating and electoral success are reinforcing. A party that has a low approval will tend to struggle electorally, but also a party that struggles electorally will lose approval from partisans who are unhappy with the approach that led to defeat. Unlike the transient bounce from above, this effect tends to last longer.
3. Similarly, a party that suddenly has electoral success will gain approval from partisans who had been discouraged.

At this point, it is difficult to discern how much of the swing in Democratic approval is due to a temporary dip in enthusiasm, and how much is due to frustrations with their election day setbacks. My bet is that there is a mixture of both at play; that the Democrats will likely recover a few points over the next couple of weeks, but their approval rate will settle somewhere a few points below the level that had been established through September.

Updated to add: Regarding the above point on response rates, please see the table on this Tom Blumer article, which shows how the response rate in polls has declined from 36% in 1997 gradually over the years to where in 2012 it was 9%. With response rates that low, if 1 in 100 Republican partisans called decides to answer when normally they wouldn't, and 1 in 100 Democratic partisans called decides not to answer when they normally would, that would be sufficient to cause a 9 point swing in the partisan composition of the sample. Weighting for demographics would somewhat mask this in the partisan identification portion, but the Republicans sampled would be over-represented by the enthused partisans and the Democrats under-represented by the discouraged partisans. In other words, the declining response rate can very well be leading to the short-term change in enthusiasm having more of a distorting effect now than a decade ago.