We have learned to track the daily twists and turns of presidential approval ratings as if they were the Dow Jones. Congressional approval questions have the same structure and employ the same language, so we naturally assume they too reflect an important reality. Thus, headlines trumpet congressional approval falling to a decade-low and Republicans now whisper their new ’08 strategy: Beat Democrats by obstructing progress to bring down congressional approval.
The underlying assumption is as straightforward as it is incorrect — congressional job approval is analogous to presidential approval.
Interpreting the meaning of responses to attitude questions is not necessarily straightforward. Results are rendered meaningful by their relationship to other responses; by comparison to the same question asked at different times or in different places, or because there is a clear relationship between answers to particular questions and some objective phenomenon in the real world.
Presidential and congressional approval questions look similar, but mean something quite different.
While presidential approval is politically consequential in its own right and predictive of electoral behavior, congressional approval is neither.
A president’s approval rating matters because it is one of his most important weapons in political combat. A president’s political power is directly proportional to his approval rating.
Congressional power, though, is unrelated to congressional popularity. Congress at 28 percent has no less influence than Congress at 40 percent.
A second attribute of presidential approval that makes it worthy of regular monitoring is its demonstrated predictive utility in forecasting election outcomes. Presidents with approval ratings over 50 percent get reelected, while those significantly below the 50 percent mark do not. Low approval ratings going into a midterm presage significant losses for the president’s party. The relationship between presidential popularity and electoral outcomes is neither simple nor linear, but it emerges clearly from merely eyeballing the data and from sophisticated statistical analyses.
By contrast, congressional approval is unrelated to election results. Some will quickly point to years like 1980, 1994 and 2006 when congressional approval sunk and the party in control suffered substantial loses. In 1980, only 25 percent approved of Congress’s performance and the majority Democrats lost 35 seats. Heading into the Democratic debacle of 1994, 23 percent approved of Congress’s performance and 54 seats were lost. Most recently, approval of Congress fell to 26 percent and the GOP lost 31 seats.
However, in trying to establish a relationship one cannot simply ignore the counter examples. In 1982, just 29 percent approved of congressional performance but the majority Democrats picked up 27 seats; and in 1992 congressional approval hit an historic low of 18 percent while the Democrats lost just 9 seats.
The blowouts have come when congressional approval is low and the same party controls both the White House and Congress. Even these few cases could well be spurious, as the presidents were rather unpopular then as well. Whatever the underlying truth, Republican strategists should note that Democrats will not face this circumstance until at least 2010.
The lesson is clear: When congressional approval is low, anything can happen. You have a better chance of predicting the stock market with changes in hemline lengths than in predicting the outcome of congressional elections using congressional approval.
While recent reports of congressional approval ratings ranging from 23 percent to 39 percent may make an interesting story, it is not a very meaningful one.
Mellman is president of The Mellman Group and has worked for Democratic candidates and causes since 1982. Current clients include the majority leaders of both the House and Senate.