We must stop obsessing about Trump voters. He got the Republican base. That is all. It sucks the white-nationalist sect of that coalition was actually happy to cast their vote for Trump, but they would have voted for Rubio too. The Republican coalition has been stable. Obsessing about them yields us nothing.
Trump, in fact, did worse with white voters than Romney overall. While he did do better with white voters without degrees than Romney, Bush scored far more votes from this group than did Trump. Trump basically maintained the Republican vote. Obsessing about these folks will not advance the goal of electing more and better Democrats.
Where we are seeing significant change is within the Democratic cohort. That is where our focus ought to be.
Voting Cohorts
Basically, the goal of a political party in electoral terms is to maintain and grow your support from various cohorts that form your coalition. Stability or growth is the name of the game. Instability and decline in any cohort within a coalition is a sign of trouble. Republican cohorts have been remarkably stable for the last 4 cycles, ours is in chaos.
To look at this I took exit polling data from How Groups Voted — The Roper Center. I wanted to do a full range of tracking on income level, age and all of the categories. Unfortunately, the data tracked changes too much from cycle to cycle over the 40 year period. For example, the high-end income level in 1976 was $20,000 and over. Today’s high end is $100,000 and over. There is no consistency to report. The same problem with age. I was going to break it between 45 and older vs younger, but many years do not have a 45-year-old cut, but rather a 50-year-old cut. That means there is no reliable way to report out those results year over year.
There are three basic polling categories I could focus on over a forty year span given the limitations in the data.
- Gender
- Male (blue line)
- Female (pink line)
- Education Level
- NonDegree Holders (yellow line)
- College Bachelor Degree Holders (black line)
- Race
- NonWhite Voters (brown line)
- White Voters (white line)
I broke these down into two categories each to define cohorts, again, largely due to reporting issues in the available data. for example, many years do not include an Asian category, while many do. That lack of consistency would skew results across categories, so that is why I break that down into white v nonwhite voters. There were similar issues with education level as well.
The Republican Coalition
Let me start with Republicans simply because I open the post pointing to the general stability of their coalition.
That is remarkably stable since 2004. It really is stable since the Reagan Revolution if you take the outlier of the Perot years out. I will say this, Democrats have long argued that Perot benefitting Clinton more is a myth. This data shows it probably is not a myth. The white cohort and the nondegree holding cohort collapsed for Republicans while Perot was a factor and then returned back to trend in the absence of Perot.
As argued above, this is remarkably stable.
The Democratic Coalition
While the Republicans have enjoyed great stability, particularly in the last 4 cycles, we have seen a great deal of turbulence.
Up until the election of Obama in 2008 Democrats enjoyed either stable or dynamic growth along all cohorts. Since 2008 we have been experiencing chaos. I will dig into that more in a moment, but our losses in certain categories are not mirrored in similar gains in the Republican cohorts, so most people leaving us from various cohorts are not going to the Republicans. That is an important thing to note as we look at what is happening.
Now, let’s take a look at what is happening with our cohorts.
Democratic Vote by Gender Trends
1976 was the only election in this 40 year period where Democrats saw more support from men than women. The story of the gender vote since has been one of stability or growth for women throughout that span. Women did show a decline in the 1980 vote, but only a slight one compared to the drop in men. So while both cohorts lost ground in 1980, we lost less ground with women voters.
The Obama Elections - both gender cohorts reached their peak level of support in the 2008 election. While the popular perception of the 2008 election was that Obama was a singularly attractive candidate, this gender data shows rapidly growing support from these cohorts for the Kerry campaign as well. In general terms, the growth in support was aggressive from the 2000 election onward. Now, that said, Obama was a uniquely gifted candidate, he was able to capitalize on the trends to secure a resounding victory for Democrats at the Presidential level being the first to clear 50% of the overall vote for decades.
Our problem stems from the fact that we saw 7 straight cycles of growth from both cohorts and now have declined since. Again, we have not lost as many women as men though.
It is important to note that women are now our largest voting cohort, even larger than our support from white voters.
Democratic Vote by Race
This category is interesting. This more confirms our suspicions than anything, although there is a surprise or two to be found.
The most important take away is that nonwhite voters represent our fastest growing cohort. That diversity is the power of our Party. The data reveals a slight problem for us though. The explosion of growth in support from non-white voters started in 1992 and only accelerated since then, well, until the 2012 election. Growth in support from nonwhite voters has finally tapered off. This cooling in the rate of growth has probably had more of an impact on the outcome of our elections than we have previously thought. It is certainly worth further discussion.
The white vote took a big hit with the election of Reagan but recovered by 1992. The white growth spiked in their support for us in the 2008 election and dropped rapidly by 2012. We maintained this cohort between 2012 and 2016 though.
Democratic vote by education level
I am not going to engage in too much commentary, I have been largely describing the trends revealed by the data. My one editorial comment is this. The changes in the education cohort represent our electoral challenge moving forward.
Bachelor Degree Holders — In 1976 those with bachelor degrees made up only 15% of the population. They are now 33% of the population. That simple growth trend represents most of our gains in this category until the election of Obama. He really ran up the score with the college educated. That near 10 million vote increase shows you where his appeal was, that is where his once in a generation status really shines through. That has come with some consequences for us however.
Non-Degree Holders — This cohort followed a similar growth pattern as the other cohorts over the last 40 years until the Obama years. Well, the exception would be the slight drop in support in this cohort during the Perot years. Our peak of support came in 2008, but we have lost 7.9 million of these voters over the last two cycles. This sort of decline is a new phenomenon for our party, one we need to consider the ramifications of thoroughly.
So what does this data show?
I really hope this starts a dialogue so I am just going to throw up a few questions to spur discussion.
- What is the biggest positive in this data for the Democratic party?
- What is the biggest negative?
- What do you think of college degree holders being the largest cohort of support for us in terms of education level?
- What accounts for the slow down in growth in nonwhite voters the last two cycles?
- Where should our focus be over the next few years?