Saturday, August 20, 2022


Recently, the average of surveys indicated that the previously substantial Republican lead in the House generic ballot test (“would you rather the U.S. House of Representatives be controlled by Republicans or Democrats?”) had shrunk to just .1%.

That has had the salutary effect of making members of Team Red nervous about the likelihood of what was supposed to be an uncontested coronation. That’s good. Competition, perceived or otherwise, tends to make people work harder.

There are, however, a couple of things about the survey results that are worth noting.

First, the averages of surveys are not methodologically sound. The wording of questions varies from survey to survey. The placement of a question in the survey can alter the responses. The day of the week or the time of day in which a question is asked or answered can affect responses. The sample to which a question is asked can be different in a number of ways.

Second, it is clear that some of the survey results in question are propaganda.

If you examine the results of the generic ballot questions in surveys included on the Real Clear Politics website, an interesting and probably not accidental pattern emerges. From the beginning of 2022 until a few days ago, there are 140 surveys that had generic ballot test results reported on RCP’s site. Of those, 107 are from surveys of registered voters. The remaining 33 are from surveys of likely voters.

Surveys of likely voters are more relevant to campaigns and election outcomes. They are better predictors of election outcomes because they include people who routinely vote, who follow the news, etc. In comparison, surveys of registered voters are sort of the lowest common denominator because they don’t account for who might actually vote.

Among the 107 surveys of registered voters, the Democrats led the generic ballot test in 46, the Republicans led in 47, and there have been 14 ties. The average result of these surveys of registered voters is about plus .3% for Republicans.

The 33 reported surveys of likely voters tell a very different story. In these surveys, the Republicans led the generic ballot test in 32 of them. The Democrats led in just one. The average result for the ballot test in these surveys of likely voters is about plus 7% for the Republicans.

In short, when you focus on results from the more precise surveys, the Republicans have a decided advantage.

That advantage is not, however, insuperable. The notion that the Republicans are going to net 65 or 75 seats this cycle is absurd and ungrounded. Democrats are outraising Republicans in most places, and when you look at the individual races, it is tough to come up with more than three dozen truly competitive races that the Republicans might win.

However, it seems reasonable to assume that the Republicans’ current advantage on the generic ballot is midway between zero and 7.5%. That indicates a majority that looks like 235 seats for the Republicans in the next Congress.

How about the Senate?

Publicly available survey work on the Senate is harder to come by, so comparisons are difficult. In Nevada, surveys of registered voters give an almost 7% advantage (on average) to Democratic Sen. Catherine Cortez Masto. At the same time, surveys of likely voters give a very small advantage (.3%, on average) to her competitor, Republican Adam Laxalt. 

Similarly, in Florida, surveys of registered voters give Republican Sen. Marco Rubio a modest 3.3% lead on average, while surveys of likely voters give him a more generous 9.5% advantage on average.

In most other cases, there are simply not enough survey results available.

The lone exception to all of this is the Senate race in Georgia. There, brand awareness is no doubt affecting the survey results, as Democratic Sen. Raphael Warnock leads by an average of 3% among surveys of likely voters, while Republican Herschel Walker leads by an average of 1% among surveys of registered voters.

At this point in the cycle — less than 90 days to election day and the close of voting — surveys of registered voters are, essentially, useless for all purposes except propaganda. Those who release them are intentionally trying to create the illusion that the election contests are closer than they actually are.

Sometimes survey results are just propaganda, and not even subtle propaganda.

• Michael McKenna, a columnist for The Washington Times, co-hosts “The Unregulated” podcast. He was most recently a deputy assistant to the president and deputy director of the Office of Legislative Affairs at the White House.

Copyright © 2022 The Washington Times, LLC.