Voter Turnout Has Nothing to do With Rising Inequality
by Benjamin Studebaker
When I talk to people about the threat rising economic inequality poses to the consumer economy, I am often told that the problem is participation. The story goes that disadvantaged groups have lower turnout, and because of this their interests are underrepresented. Bernie Sanders plans to win support for his egalitarian agenda through a “political revolution”, but all he means by that phrase is exceptionally high voter turnout:
…if Bernie Sanders becomes President of the United States, it will mean that there will be a huge increase in voter turnout. If there is a huge increase in voter turnout, our Republican colleagues may not be running the Senate or the House. So my life will be made a little bit easier. As I mentioned earlier, Louise, Republicans do well when people don’t vote. For me to get elected, we’re going to have to have a huge increase in voter turnout, and that will carry in a lot of other people in the Congress and Senate
The turnout story is especially popular among left wingers in the United States, because the US has unusually low voter turnout for a rich democracy. But the refrain is heard abroad as well–left wingers in Britain think that Jeremy Corbyn can win by raising turnout. There are right wingers who make similar arguments–Ted Cruz claims that republicans lost in 2012 because of low voter turnout among evangelical Christians. It’s a seductive argument, because we all like to believe that our own beliefs are common sense and that there’s a silent majority out there who agree with us. Unfortunately for all those who make turnout arguments, the data indicates that they are bunk.
The trouble with a lot of voter turnout research is that it myopically focuses on American politics and it makes crucial false assumptions. In 2014, Pew conducted research on the political beliefs of non-voters in America and found that despite major differences in demography (non-voters tend to be non-white, young, and poor), non-voters held political beliefs that were remarkably similar to those of voters. They were only slightly more sympathetic to Obama and the democrats than voters:
And they were only slightly more likely to approve of Obamacare, immigration reform, and welfare spending:
But at other times, American non-voters seemed to pull further left. When Pew ran a similar survey in 2010, it found a bit more distance between voters and non-voters, with non-voters pulling closer to the democrats on every issue except gay marriage:
The trouble with this research is that by focusing on non-voters’ issue stances, researchers implicitly assume if non-voters were to vote, they would vote in accordance with those issue stances. One the key things the 2010 Pew survey indicated is that as a group, non-voters don’t pay attention to what’s going on:
We already know that a lot of voters don’t fully understand the effects of the policies proposed by various candidates and parties. Bernie Sanders is considered a profligate spender while the republican candidates are considered fiscal conservatives, even though the republican tax plans widen the deficit while Sanders’ proposed spending programs are remarkably affordable. Yet voters who are concerned about balancing budgets are far more likely to support republicans than they are Bernie Sanders. They’re judging based on what everyone is telling instead of doing the research necessary for a well-informed opinion. There are many other issues like this where people’s stated policy preferences are not reflected in their political choices. Hillary Clinton leads Bernie Sanders among women and blacks even though Sanders has a more robust record of supporting policies that help both of these groups, especially economically. If voters can be deeply uninformed about which candidates and policies will really accomplish their objectives, non-voters are surely much worse off. This is reflected in the reality that non-voters are far more likely to support Donald Trump, according to recent research:
Another turnout challenge for Mr. Trump is that he commands the support of many people who are unlikely to vote. Civis found him winning 40 percent of the vote among those it gave less than a 20 percent chance of participating in the general election — let alone in the primary. He held 29 percent among those who had greater than an 80 percent chance of voting in the November election.
I’ve been getting increasingly suspicious of turnout arguments, which always and everywhere assume that non-voters will proceed to vote rationally for whichever candidate the person making the turnout argument thinks is best for them. To sort out whether these arguments are really any good, we need to look outside of individual national contexts and get a real sense for how turnout affects long-term policy outcomes in rich democracies. To that end, I took a dozen wealthy democracies and compared their voter turnout rates to the size of the increase in the top 1%’s share of income since that country’s respective post-war low. My turnout data on the last election is from Pew while the averages come from the International Institute for Democracy and Electoral Assistance. My inequality data is from the World Wealth and Income Database. What I found was striking–there is no connection whatsoever between participation and the distributive outcome. This is true both for the most recent national election:
And it is true when we average participation rates together for all elections in the second half of the 20th century:
The UK has seen nearly the same increase in inequality as the US despite an average participation rate that is more than 20 points higher. Countries like Australia have higher turnout than France or Japan yet see inequality increase far more rapidly. Italy has the highest average turnout, but performs worse than France, which has one of the lowest turnout rates and one of the best inequality marks. We can find examples of countries with very high turnout and low inequality growth (Netherlands & Denmark), very low turnout and low inequality growth (France & Japan), very high turnout and high inequality growth (Australia), and very low turnout and high inequality growth (USA).
Even if non-voters have a greater stated preference for equality, even if they are more likely to come from disadvantaged or disempowered groups, we have no reason to believe that getting non-voters to vote will result in policies that help these groups. This reflects an important reality that is not widely understood–the vote is only an effective means of defending a social group’s interests if that group is able to consistently select the parties, candidates, and policies that advance that group’s interests, and most groups are rubbish at this. Disadvantaged voters are terrible at it because they are especially poorly educated and don’t pay attention. When they vote, they are as likely to be manipulated into supporting people like Trump or UKIP’s Nigel Farage as they are to be used effectively in building left-wing coalitions to fix the distribution. They can also be readily and easily used by Clintons and Blairs, who pretend to care about about inequality while actually widening it rapidly, damaging the consumer base and making their economies overly reliant on finance, leading to economic instability and underconsumption:
The problem is not that people don’t vote–it’s that when they do vote, they vote badly. If your preferred candidate is waiting for a “political revolution” where some silent majority of egalitarians rises from the electoral mists as if summoned by some unholy necromancer, it may be a long wait.
Alas, what none of the polls are able to measure is laziness, ignorance and stupidity.
Typo in the fourth paragraph: “non-voters held political beliefs that were remarkably similar to those of non-voters”. Either that, or… yes, I expect so.
Good catch! Thanks for that.
A long, long time ago when I was young and naive I became very active in behalf of a Midwest congressional candidate who happened to be a political science professor. I met with literally thousands of voters in the district, some well educated and others not so much. In those years there were few minority voters in the district and the age for voting in federal elections was still 21.
One particularly shocking thing to me was how many registered voters — across the board — didn’t know the difference between a state legislator and a U.S. congressman. Mention the candidate I was supporting “for Congress” and more than half the time the voters assumed he was running to “go to [the state capital city].” Mention a second long-time state senator, state rep, or even a county office-holder and the voters who had just gotten the first question wrong grew wary and assumed he served in Washington.
Many more than half of the several thousand voters I talked to over the six months or so I worked on the campaign had no bleeping idea who, what, or where the office was for any incumbent or aspirant.
Another true anecdote (related by a friend) also supports your statistics. Some time after President Ronald Reagan announced his approval for the “Pickle Amendment” (named after a retrograde Democratic congressman from Louisiana who sponsored legislation intended to effect a deep reduction in the percentage of Social Security Disability beneficiaries) and about a month after Reagan called on Congress to eliminate federally-funded Legal Services for the poor, an older, very sick, rural man visited the local Legal Aid office. There he was interviewed by a friend of mine.
The man was being threatened with revocation of his SSD benefits — consistent with the Pickle Amendment. He had no money, so he sought legal help from the Legal Services office. He got the help he wanted (and eventually prevailed in court). After that first visit, he left the office and climbed into his old rattle-trap pick-up truck. My friend followed him out to his vehicle. On the back bumper and again on the back window were two old bumper stickers touting: “Ronald Reagan for President.”
My friend could do nothing but shake his head in disbelief.
Very good post as always. There is a typo in this sentence: “The trouble with a lot of voter turnout research is that it myopically focuses on American politics and it makes a crucial false assumptions.”
Good catch! Thanks for that.
It’s not just about voter turnout. It’s about turnout of particular kinds of voters.
The turnout issue is about more than numbers. It’s about the who that the numbers turn out for and the who that stays home. While the politics of the non-voters may be vaguely similar and equally distributed, the Republicans are highly unlikely to turn out those non-voters who otherwise might vote Republican.
The most important voter turnout that matters is the turnout of currently disaffected voters — people who are tired of politics as usual, people who have voted in the past and have become disillusioned with the existing two-party system. Second in significance are the new voters — the young whose imaginations get captured by a particular message.
Both turned out for Obama in unprecedented numbers in 2008 and 2012. At the same time a lot of disaffected Republicans stayed home because they couldn’t stand Romney.
We now have a new generation of disaffected voters who are unhappy with Obama’s failure to live up to the vision that captured them in the first place. That is, the young voters who turned out for Obama and the Obama voters who enthusiastically embraced the fresh air he brought to the scene. And then there are the people who couldn’t vote in the last two elections but are old enough now.
Hillary isn’t capturing those voters. Bernie is.
The other side of the coin is disaffected and new Republican voters. There are a lot of Republicans who are thoroughly disgusted with the radical right wing, racist/sexist version of their Party.
There are still registered Republicans who identify with the Party that you find in the Republican platforms of 1968 and 1972. They haven’t yet traded their R card for a D card, but they are still there. Those voters They are likely to either stay home or to vote Democratic. Interestingly, those voters seem to be more attracted to Bernie than to Hillary. Even some Evangelical leaders have endorsed Bernie.
As for the “laziness” issue: for a broad spectrum of voters it’s not about laziness as much as it is about being disaffected. While there are certainly plenty of apathetic voters, many of the people who stay home on election day have lived with decades of being disappointed with the electoral process. In their view, neither party gives a damn about them, and no matter who is elected, nothing changes.
The apathetic voters won’t turn out no matter what. The disaffected voters will turn out if they can be convinced that a candidate will make a difference in their lives and the lives of loved ones.
Let me illustrate that with some hypothetical numbers.
Take a hypothetical 1 million stay-at-home voters. Split them evenly down the middle between Republican-leaning and Democrat-leaning. Candidate A in Party 1 has a breath of fresh air campaign that attracts 250,000 of them, while Candidate B in Party 2 turns people off so much that all of their allotted 500,000 stay home. So that is a net gain of 250,000 votes for Candidate A and no gain for Candidate B.
THAT is how turnout makes a difference.
“I am often told that the problem is participation.”
This statement seems to be the fundamental premise of your argument. There are two issues that I think your article fails to defend, whether or not correct.
When you refer to participation, you immediately reduce that to voter turnout; yet when referring to participation one also can intend to mean other forms of participation, such as informing oneself and actively starting and joining political discussions. The source you cite to this from the Daily KOs was unavailable to me. Based solely on your block quote, I think you uncharitably pigeonhole his point, which is that if voters turn out in the majority for him, they will likely turn out in the majority for Congressional candidates who also support similar policies. The entire criticism of voter turnout seems to be precisely that “mere voting” isn’t enough, participation requires a more motivated and engaged electorate.
The second issue dovetails the first issue into your conclusion. The problem isn’t a lack of voting, it’s that voters frequently vote against their self-interest (I assume that’s what you mean by vote-badly). My concern isn’t that you seem to blaming the voters, but rather that you fail to address the clear disparity between expectation of the voters and effect of policies. It seems to me any politician capable of winning an election ought to have some idea of the effects excepted by his or her electorate. Either you are suggesting the voters are incapable of meaningful participation by their own failure, or are prevented from meaningful participation by the deceptive tactics of candidates. Shouldn’t the “free press” that was so necessary to our democracy bear some responsibility for not making readily available the information necessary for meaningful participation?
Quality of the participation is an entirely separate argument from quantity. I absolutely agree that false consciousness is a major problem, but in this particular post I am discussing the claim that the quantity of participation is not high enough, not the claim that the quality is poor.
“The trouble with this research is that by focusing on non-voters’ issue stances, researchers implicitly assume if non-voters were to vote, they would vote in accordance with those issue stances. One [of] the key things the 2010 Pew survey indicated is that as a group, non-voters don’t pay attention to what’s going on:”
Missed an [of] ^^
Great post, but you screwed up my Call-for-Action speech topic in which I was going to argue for greater voter turnout to counteract our shift toward oligarchy! 🙂
It seems to me that you are making 2 large claims:
1) Mobilization of non-voter groups are unlikely to effect inequality
2) Non-voter groups do not vote rationally
Yes they are definitely related, however I feel that you are making large sweeping generalizations about non-voter groups that may not be accurate.
I think it makes more sense to discuss your second claim.
To support this claim, you utilize survey results to reflect the preferences of non-voter to which you claim are inconsistent with their voting patterns. First, I think its fairly obvious that survey results are not reliable for a number of reasons (questions, targets, etc). In particular, survey results do not indicate the magnitude of the preferences of non-voters. Yes non-voters may indicate that they support more welfare, however that does not mean they believe that is their highest priority. To suggest that non-voters based on survey results vote irrationally or “badly” because they do not vote based on their perceived economic interests, I think is huge mistake. Yes it is somewhat strange that many supporters of trump are in tension with their economic interests, however to apply a generalization that this tension results in a contradiction where voters are irrational is incorrect. Non-voters may still be voting rationally if they value for example Trump’s leadership or vision of the state more than they do about marginal differences between economic policies. Furthermore, recognizing that especially amongst non-voter groups general attitudes in whether presidents are able to carry out their economic plans is justifiably low, it may very well be rational for voters to vote on other points.
To discuss your 1st claim now concerning how non-voter groups if they were to vote would not effect inequality, I feel is an assertion that you cannot make based on your reasons or evidence. For the sake of argument, let us grant that non-voter groups may not vote rationally, or may vote badly, from this premise you cannot reach your conclusion that therefore inequality will not be effected. At best, you are able to reasonably claim that it may or may not effect inequality, either for worse or better – which does not really help us. The evidence you provide of the charts that you claim show no correlation between voter turnout and inequality has a few problems. First, the charts take the average of a 50 year span of voter turn outs which does not capture individual terms. For certain years voter turn outs may have very well have inequality gaps. Furthermore, your evidence does not indicate on a country level how during the 50 years voter turnout could have effected inequality. In particular I could point to a number of papers that utilize actual econometric analysis that shows the effects of voter turnouts on economic inequality (here is one of them that also does a good job of reflecting the current economic literature on this topic (Economic Inequality and Electoral Participation. A Cross-Country Evaluation Antonio M. Jaime-Castillo). Although the current economic literature shows differing results (positive, any, negative) on how voter turnout effects inequality, this supports only the claim that we are not sure of the relationship between the two.
In general I enjoyed reading your piece, it was very well written. However, I feel for the claims you are asserting not enough support exists.
Voters may be voting for non-economic reasons, but this would only make them rational in the sense that their vote tracks their preferences. It would not justify those preferences as being in line with their objective interests. Regardless of whether voters care about their economic interests, the claim is that they are not voting in line with them.
Individual terms are irrelevant because policies that affect the distribution of wealth often have very long lag times. Using a long period is the only effective measure because it allows the cumulative effect of turnout on the distribution over the whole period to be shown. It is simply not possible to do an interesting or meaningful one-country study over short terms when the consequences of policies may not show up in the distribution for a decade or more.
The research you cite looks at the effects of inequality on turnout rather than the other way around. It is engaging with an entirely separate literature about the effects of income on participation among different voting groups. The study also chooses to use income ratios rather than top 1% share, which makes it impossible to see inequalities that are concentrated at the very top of the distribution. These are the inequalities that contemporary writers in the field view as most threatening to the economic and political order.