The Last Two Weeks of Polls Have Been Great for Republicans. Do They Signal a Shift?

Last month it seemed that Democrats might ride a giant tsunami to control of the House and Senate. Now, some are wondering whether there’s a Democratic wave at all.

The Democratic advantage on the generic congressional ballot, which asks people whether they’ll vote for Democrats or Republicans for Congress, has dwindled since the heart of the tax debate in December. Then, nearly all surveys put Republicans behind by double digits. Now, poll averages put the Democratic lead at only around six or seven percentage points.

The question isn’t really whether Republican standing has improved recently. It has. The question is whether anyone should care: Is it just one of many blips and bumps along the road, or does it say something meaningful about the midterm elections?

The short answer: Check back in a month.

The shift hasn’t lasted long enough to merit a reassessment of the national political environment. But there are reasons to think it could.

It can be hard to remember, but we’re still nine months from the election. In a presidential election year, the New Hampshire primary might be coming up next Tuesday. So it is quite early to care about every little bump on the congressional generic ballot trend line.

And, to some extent, the Republican gains look like another little bump. After last month’s Democratic surge, it was reasonable to expect the polls would revert toward the longer-term mean, which had been something closer to an eight- or nine-point Democratic edge.

But the last two weeks of polls have gone further than a reversion to the mean. They’re arguably the best two weeks of polls for Republicans since the failure of the Senate health care bill in July. A highly sensitive poll average — like the FiveThirtyEight tracker — might put the Democratic lead down to roughly six points, basically the lowest level since the spring.

One can imagine any number of explanations, but perhaps the best is changing public opinion of the Republican tax plan.

In December, polls showed the plan was deeply unpopular. This was when the bill was grinding its way to passage in Congress. Its opponents were relentlessly campaigning against it, while the bill’s supporters were busy negotiating — and perpetually dissatisfied with its latest form — and didn’t defend it in a meaningful way. Most voters thought they would see a tax hike, even though they were probably going to get a tax cut.

Today, the politics of the issue are very different. The attacks on the bill have relented now that it has passed. A steady stream of good economic news, including announcements of bonuses and raises that some corporations are attributing to the tax cut, have amounted to a substantial if belated public relations campaign for the bill. It is also possible that voters have realized they’ve gotten a tax cut.

The bill isn’t exactly staggeringly popular, according to more recent polls that still show that more Americans disapprove of the bill than support it. But it’s polling a lot better than it did a month ago and even fairly well by the standards of President Trump’s approval rating.

Regardless of how it polls over all, there can be little doubt that many Republicans, who might have been dissatisfied with the accomplishments of the Republican Congress, have now gotten one of their biggest wishes.

Does that add up to a lasting change in the national political environment? Who can say. What can be said, fairly confidently, is that two weeks of good polling for the Republicans are not enough to declare that such a shift has taken place.

Take a look at two ways of calculating the generic ballot polling average. Both account for the biases of various pollsters, and give more weight to higher-quality and less frequent pollsters. In light gray is a sensitive estimate — the sort of polling average you would use heading into Election Day. It doesn’t care much about polling from a few weeks ago. Wednesday’s Monmouth poll, which showed Democrats ahead by just two points, is far and away the most important single data point in its estimate.

In dark purple is a less sensitive estimate. It’s as sensitive as the Upshot Senate model from 2014 would have been nine months from a congressional election. It’s not sensitive at this stage for a simple reason: Bumps and blips don’t tell you too much this far from an election.

At the moment, the two estimates tell very different stories about the generic ballot. The longer-term estimate is essentially right in line with the average since late July, when health care reform failed in the Senate. The short-term average, on the other hand, is as low as it has been since the F.B.I. director Jim Comey was fired in May.

As an empirical matter, the longer-term estimate is generally the better one to follow this far from an election. And looking back on how these estimates have played out this past year, it’s fair to say that the empirical case for the less sensitive estimate has generally held. Most of the seemingly big shifts have proved to be ephemeral. Every time there has been a big gap between the two estimators, the polls have reverted toward the longer-term average.

You could go even further and argue that even the less sensitive estimate is too sensitive, and claim that the Democratic advantage on the generic ballot probably didn’t materially change between June and November — other than perhaps a subtle and steady trend toward the Democrats.

There are certainly reasonable arguments for why strong economic growth and a more popular G.O.P. legislative agenda might lead to a meaningful shift toward Republicans. But it’s too early for that conclusion: We’ll probably have a better idea in March.

Even if the current, diminished Democratic advantage does last for the next month, it won’t mean the Republicans are in the clear. Democrats would still have a very real chance to take back the House with a six-point lead on the generic ballot, and there’s a tendency for the party out of power to make gains on that ballot as midterm elections approach.

The fight for control of the House would probably be considered a tossup, whether the Democrats held a six-point or eight-point lead that far from the midterms.