Postscript to “Losers Win”
September 20, 2020
I am grateful to everyone who read, disseminated, and responded to my essay. Critical responses have generally fallen into two categories: (1) Nice analysis, but what do we do with it? and (2) Opinion polls are not accurate measures of people’s real beliefs; elections are.
I want to address these two criticisms here.
First, what do we do? The only comprehensive way to correct the unfairness of our current political system, and make the system truly democratic, would be to scrap our Constitution and start over. Other countries have more modern and democratic constitutions. The American Constitution is an 18th-century document created by men who did not fully trust democracy and therefore included structures and provisions (such as the electoral college) to prevent the people from exerting too much direct power over politics and politicians. Since then, Congress has addressed some of those initial flaws (by, for instance, giving women and formerly enslaved people the right to vote). However, the document is a patchwork of amendments and compromises. The entire structure of the Constitution is flawed in terms of democracy, allowing for minority rule. We certainly want to protect and retain the rights of the minority, but we don’t want minority rule, which is dangerous. From the beginning, the U.S. Constitution was a compromise with the small states (termed “The Great Compromise”) to get them to join the new union. Madison’s original plan involved representational government in both the House and the Senate. We are still dealing with the fallout of that initial compromise, its problems exaggerated now for reasons I tried to articulate in my essay.
Of course, replacing the Constitution is not realistic, and there is no viable movement in the U.S. to do so. In fact, I’ve never heard anyone seriously suggest it, although I suppose I just did. But there are some more modest solutions that would advance the patchwork approach the U.S. has already taken to addressing Constitutional flaws. First, we could end gerrymandering in all states and replace it with non-partisan districting systems. Second, we could replace the electoral college with a national popular vote, which may not require a Constitutional amendment (click here and here to learn how we might do so). Third, Congress could award statehood to Washington, D.C. and Puerto Rico, creating four more senators from small states that lean Democrat. And fourth, Democrats could move out of urban areas and large Democratic states (like California) and into suburbs, rural areas, swing states, and less populated states. I’m serious about number four: In fact, Democratic migration would do more to create election fairness than anything other than replacing the Constitution itself. These patchwork solutions would not fully address the inherent problem with the Senate, nor would they address Republican obstruction, destructiveness, and Machiavellianism, but they would tilt the scale a bit more toward a fair balance and more democratic (small ‘d’) election outcomes.
The second response I received was that polls are inaccurate measures of people’s real opinions. There’s no problem with the Constitution. Rather people lie to pollsters, as evidenced by the 2016 election. So we should look to elections themselves to find out what people really believe.
This is a serious substantive objection, but it’s wrong.
Polling is very accurate, including the polls for the 2016 election. The final estimate on the national popular vote for president, averaged by RealClear Politics the day before the election, had Clinton ahead of Trump 3.3 points, very near the actual popular vote gap of 2.1 points and well within the margin of error. The pundits got it wrong, but polls were accurate within 1.2 points; that’s astonishingly prescient. Even polls that had Clinton ahead in swing states that went for Trump were accurate within the margin of error for the poll. The polls predicted the wrong outcome only because they predicted a close race, and in close races pundits may neglect to take into account the margin of error. Also, there is no evidence that people expressing their opinions to pollsters are lying. It’s not as though shy Trump voters told pollsters they’d vote for Clinton; they responded “undecided” or refused to respond. When 71 percent of Americans tell Gallup they support Roe v. Wade, and 23 percent say they want it overturned, there’s no reason to think people are lying. We just don’t know what the “undecided” 6 percent believe.
The claim that people lie to pollsters relies on the same false assumption that my essay refutes: that election results, not Gallup polls, tell us people’s real beliefs. American elections do not tell us people’s real beliefs because of distortions in the election system. If you tally the actual votes (not the outcomes) in races for the House, the Senate, the Presidency, or all of the state houses, you find that most Americans support the Democratic Party, typically by fairly large margins, just as polls would predict. The current Senate — in which Republicans hold 53 percent of the seats, despite the fact that Democratic Senators received 15 million more votes — is just an extreme example of the anti-democratic inequities in our American election system. Throughout the system, gerrymandering, demographics, an old-fashioned Constitution, and an unscrupulous Republican Party subvert the majority and enable Republican victories that the principles of democracy say the party does not deserve. Our 18th-century Constitution, a compromise from the start, is too old to handle modern politics in any way that resembles a real modern democracy.
