The Democratic Party is losing its moderation and its middle age

I looked at the Democratic Party presidential nominating contest exit or entrance polls for 2016, 2008 and 2004. (As Monica knows, I am extra, but in a good way.) I found 27 states*’ exit or entrance polls for the first two of those years. The averages:


18–29 turnout increased 1.7 points.

30–44 decreased 1.6 points.

45–64 decreased 3.1 points.

65+ increased 3.4 points.

Very liberal increased 8.3 points.

Somewhat liberal increased 6.8 points.

Moderate decreased 8.9 points.

Conservative decreased 6.3 points.

Given that no one age demographic moved anywhere close to how much the ideology demographics moved, we can’t simply say that fewer 45–64 voters or more 65+ voters moved the needle for moderate voters. And in fact, several states show that point even more clearly. Check Iowa:

Iowa is one of two states (hi, Ohio!) in which 18–29 turnout decreased for the years in which we have both sets of data.

Now, conventional wisdom — which is to say “wisdom” from a rich, straight, white male political pundit who gets paid whether he is right or badly, badly, shamefully wrong — might suggest that a decrease in 18–29 turnout and a sharp increase in 65+ turnout would lead to more right-wing numbers, right?

You could not be more wrong if you were an Iraq War cheerleader now trying to get Bernie to apologize for opposing that bad war from the start. See, we have a sharp trend in the liberal direction — TWELVE points of moderates going away — and that, to me, says Iowa Democrats got more liberal.

How else do you explain Iowa being 10 points more very liberal and 12 points less moderate? Everyone, including senior citizens, sees the world as needing more change from people like Bernie (and us).

But Iowa is basically always wrong when asked for its opinion on presidential politics, so I shouldn’t use it to prove a point about people who have a clue. Unfortunately, my all-stars are either unavailable (Alaska’s, D.C.’s and Wyoming’s exit polls are hiding from me) or … biased (hello, Vermont!). Michigan is a bit of a suspect customer thanks to 2008’s shenanigans, and North Dakota is Sir Not Appearing in This Analysis owing to, again, missing exit polls from 2016. (If anyone knows of a repository of such things, please do holler.)

So we turn instead to the relatively reliable Carolinas, plus Oklahoma.

These are not states that anyone thinks of when asked, “Where would you go to find a bunch of liberal Democrats?”

And South Carolina is, indeed, not liberal:

You might see that 23 percent very liberal and think, “Holy heck that’s a lot of liberals!” And compared to West Virginia (17), yeah, South Carolina is very liberal. But of the states with data, the average very liberal figure in 2016 was 25.69, so 23 is on the conservative side.

South Carolina’s demographic movement helps illustrate my suspicion that many voters are getting more liberal and that age isn’t necessarily a factor. South Carolina gained 7 points of very liberal voter from 2004 to 2008 (13 to 20), and the 18–29 category’s 5-point gain in that time (9 percent to 14 percent) doesn’t explain the shift entirely. And in case you’re thinking the 30–44 crowd is responsible for the rest of that shift, look at the 30–44 crowd shrinking from 2008–2016, the 18–29 crowd growing only a point and the very liberal vote continuing to grow. If the 30–44 crowd were driving a liberal shift, their lesser presence in 2016 compared with 2008 would lead to fewer liberals in 2016, not more.

Instead, I conclude that moderates are moving left, as are conservatives. Between 2008 and 2016, the conservative Democratic presidential primary electorate shrank from 13.32 points to just 7.04. That might represent people crossing over to vote for Trump, but if that happened, the redistribution of percentage points would be more even.

Now, our next state, North Carolina:

Between 2008 and 2016, the 18–29 crowd increases by 4 points, the 45–64 crowd decreases by 3 points, and very liberal goes up 10 points and conservative drops 12. Age alone doesn’t explain anything. Instead, Democrats are getting more liberal. And if it’s conservative Democrats crossing over and voting for Trump, it’s not just that factor — or we’d see the 12 points of conservative movement from 2008 to 2016 reflected in more even distribution than 10 very liberal/2 somewhat liberal/2 conservative. Even if moderates cross over too, a 10/4 split of increase in very liberal and somewhat liberal tells a different statistical tale.

One more example. Oklahoma:

This ought to fascinate you.

Oklahoma got far younger in 2008 than it was in 2004–3 points to 18–24, 3 to 30–44, 6 from 65+ — and got more liberal. One might argue that the movement left corresponded neatly: 4 points from moderates and 1 from conservatives were deposited largely in somewhat liberal.

In 2016, all statistical hell broke loose: The old folks returned, up 4 points. 45–64 lost 7, and 18–29 gained 3.

Result: Very liberal up 8. Moderates down 8. Did they go to Trump? Again, we’d see closer to uniform redistribution ideologically and chronologically. Instead, we get an age-agnostic ideological shift.

It’s almost like grandparents and their grandkids both wanted to help fulfill Bernie’s vision for America. He did win the state, after all.

And speaking of grandparents …they’re plenty liberal. You cannot rely on them to keep this country from marching toward Medicare for All.

Maryland’s elderly voters are a great example of this.

Now, I previously cautioned you against listening to anything Maryland says about who should be the next president, since the state has the worst record of any state dating back to 1976 (and further, if anyone cares). But Maryland’s elderly Democrats are a great example of what I’m talking about. Here’s why:

Between 2004 and 2008, the state got far younger — shedding 8 points of 65+ and 3 points of 45–64 and gaining 6 of 18–29 and 5 of 30–44. And it got, y’know, a little more liberal. Conservative fell 2 points, somewhat liberal fell a point, and very liberal rose 2. So while the voter age distribution changed, the sentiment did not.

Then 2016 happened, and while the 18–29 and 30–44 crowds stayed put, 45–64 ceded 4 points to 65+. Maybe people just got older?

Whatever happened, people sure got more liberal: 5 points to very liberal, 5 points to somewhat liberal, 4 points from moderate, 5 points from conservative.

(If you were going to make the case that a state didn’t get more liberal but just lost its conservative Democrats to the Republican Party’s presidential primary, Maryland would be a significant entry in your data-based argument. It’s still a hard case to make, but it isn’t unreasonable.)

Ready for a fun one?

West Virginia, at 17, is the state with the smallest very liberal presidential electorate of the states whose 2016 exit or entrance polls I could find. West Virginia is roughly twice as conservative as the average state with accessible data. Only Arkansas and Oklahoma were more conservative by my metrics.

And West Virginia couldn’t get enough of Bernie.

He beat Hillary among men.

He beat Hillary among women.

He beat Hillary among young adults.

He beat her among old adults.

He beat her in the 17–29, 30–39, 40–49, 50–64 and 65+ demographics.

He beat her among the very poor and the very rich.

He smashed her among independents. (She beat him by 4 points among Democrats.)

So this notion that a conservative state won’t vote for Bernie/socialism is nonfactual. Alaska, Indiana, Montana, Nebraska, North Dakota, Oklahoma, West Virginia and Wyoming voted for Bernie/socialism. The only reason I’m not including more conservative states in this analysis is that, again, I couldn’t find exit/entrance poll data for them.

But getting back to the 65+ voter, the interesting thing about West Virginia is that from 2008 to 2016, 18–29 grew by just 1 point and 30–44 stayed put. But 45–64 shrank by 5 points, 65+ grew by 3, and very/somewhat liberal grew as much as moderate/conservative shrank.

What I see is that West Virginia’s elderly population made the state more liberal.

And again: If it were just 45–65 defecting, the other age brackets would have gained more evenly (or at all). If it were just the conservatives leaving, every ideology bracket would have gained.

Old Democrats are becoming more liberal. Perhaps they want their grandkids — people who are my age or younger — to have a better future.

If the Democratic Party is to beat through the thickets of lobbyists, corporate psychopaths and people who are scared of having hard conversations with abusive relatives, it will need a legion of elderly people at the front.

People who have lived through the “Social Security is socialism!” scare.

People who remember the “Medicare is socialism!” scare.

People who have seen our social safety net be enough and have seen people flourish as a result.

And as the data show, those people are in every state, and they are already leading us.

(*The states: Alabama, Connecticut, Georgia, Illinois, Indiana, Iowa, Maryland, Massachusetts, Michigan, Mississippi, Missouri, Nevada, New Hampshire, New York, North Carolina, Ohio, Oklahoma, Pennsylvania, South Carolina, Tennessee, Texas, Vermont, Virginia, West Virginia, Wisconsin.)

I write mostly data-driven stuff.