Mellman: Meeting in Miami

June’s Democratic debates seem to bear out the conclusion offered in my last column — primary debates can have more impact than general election confrontations.

Pundits and polls agree that Sen. Kamala Harris (D-Calif.) had an excellent night. The commentariat didn’t need any numbers to declare her the winner, but here the numbers back them up. 

{mosads}I look at them in three simple ways. 

First, compare the average vote for each candidate before and after the debate. By that measure, Harris added 8 points to her total. 

Second, several pollsters had surveys shortly before and just after the debates. Examining just those differences factors out house effects — internal practices that can affect results.

By that reckoning, Harris gained a very similar 9 points. 

Finally, Morning Consult and FiveThirtyEight together conducted a large sample survey just before and just after the debates. In that poll, Harris picked up 8.9 points.

So, we can be reasonably certain the debates and their aftermath gave Harris a meaningful boost of 8 to 9 points.

The other generally accepted conclusion was that the debate didn’t help former Vice President Joe Biden.

Two of my three approaches suggest his support fell by 5 points. The Morning Consult-FiveThirtyEight survey suggests he lost a slightly larger 7.5 points.

His favorable ratings dropped by 2 points, one of only two candidates who saw a decline on that indicator.

That said, what got lost in the flood of commentary is the fact that despite Biden’s falloff and Harris’ surge, the former vice president remains in strong position.

That’s what defines a front-runner—the ability to take some hits and stay on top. 

On average, post-debate, Biden still enjoys a double-digit lead, with Harris now holding second place, less than 2 points ahead of Sens. Bernie Sanders (I-Vt.) and Elizabeth Warren (D-Mass.).

The Morning Consult survey gives Biden a slightly larger 15-point lead.

Is Biden destined to remain No. 1? I have no idea, and neither does anyone else. These data cannot tell us that.

There was one other point of consensus among commentators—that Sanders also had a rough night. 

Here the evidence is modest. Two of my methods of comparison show a loss of 2.4 to 3.4 points. The Morning Consult-FiveThirtyEight poll recorded a half-point gain for the Vermont senator. 

With other candidates, the pundits and the data diverge.

For example, Warren was also treated as a big winner. 

Yet, on average, her vote total was about a point lower after the debate than before it. Not a real fall, but certainly no big gain. Morning Consult-FiveThirtyEight indicated a small Warren gain and her favorables rose by almost 10 points. 

In fairness, she appeared to be moving up before the debate and seems ensconced in the top four, at least for now.

Moreover, immediately following the first debate, her vote jumped 5 points, indicating a very good performance, but her gains dissipated after the drama of the second debate.

Mayor Pete Buttigieg of South Bend, Ind., too was judged a big winner by pundits, a conclusion that finds no support in the data. All three comparisons suggest his support actually declined ever-so-slightly.

Former Housing and Urban Development Secretary Julián Castro also got credit for a great performance, and indeed, his favorables skyrocketed by nearly 20 points, the biggest gain than for any candidate.

But that was not translated into additional support, with movement of less than a point in both directions, depending on which comparison one examines. 

What does all this tell us about future debates?

First, insta-analysts should be careful about picking winners. Sometimes they are right, sometimes they’re wrong.

Second, with so many candidates it’s very difficult for more than one to really break through. More than one big winner seems unlikely.

Third, there’s still a good deal of fluidity in this race.

Finally, a great performance the first night can be wiped away by bigger drama, and bigger fish, on the second. 

Mellman is president of The Mellman Group and has helped elect 30 U.S. senators, 12 governors and dozens of House members. Mellman served as pollster to Senate Democratic leaders for over 20 years and as president of the American Association of Political Consultants.