Swings and round-abouts

A couple of thoughts on The Congressional Map Has A Record-Setting Bias Against Democrats by David Wasserman. I like FiveThirtyEight as rationalists striving to use logic and evidence to arrive at greater clarity about the world, trends, and patterns around us.

For all that that is their goal, it is also interesting to see the errors in the process, the occasional blindness to context, the unstated assumptions.

In this article, Wasserman is highlighting the challenges faced by Democrats in the 2018 election cycle. No argument with the general direction of the article and its main conclusions.

When Democrats think about their party’s problems on the political map, they tend to think of President Trump’s ability to win the White House despite losing the popular vote and Republicans’ potent efforts to gerrymander congressional districts. But their problems extend beyond the Electoral College and the House: The Senate hasn’t had such a strong pro-GOP bias since the ratification of direct Senate elections in 1913.

Even if Democrats were to win every single 2018 House and Senate race for seats representing places that Hillary Clinton won or that Trump won by less than 3 percentage points — a pretty good midterm by historical standards — they could still fall short of the House majority and lose five Senate seats.

This is partly attributable to the nature of House districts: GOP gerrymandering and Democratic voters’ clustering in urban districts has moved the median House seat well to the right of the nation. Part of it is bad timing. Democrats have been cursed by a terrible Senate map in 2018: They must defend 25 of their 48 seats while Republicans must defend just eight of their 52.

It is the last paragraph I find interesting. There is a passivity to it that is odd. Things have happened to Democrats rather than Democrats have done things that to put themselves in a weak position. The crux is "GOP gerrymandering and Democratic voters’ clustering in urban districts."

I see three problems in that nine-word sentence fragment. The first problem is that gerrymandering is a bipartisan activity of long standing. Our current electoral boundaries are entirely a product of Democratic and Republican gerrymandering. Granted that increasing Republican electoral success has made their gerrymandering more prevalent, but that begs the question. What was it that Republicans did to win so much territory when Democrats were dominant and the electorally gerrymandered boundaries were set by the Democrats? Gerrymandering was not the root cause of Republican success, it was some set of policies or strategies or leadership. By putting his point in the passive voice, Wasserman directs attention away from real root causes.

The second problem in the sentence fragment is that there is a reasonable body of academic research which suggests that gerrymandering is not near the boogeyman as commonly believed. Yes, it makes a difference at the margin and therefore is relevant in very tight races but overall, at the national level, it rarely affects the national outcome. My recollection is that only a small handful (single digits) of House seats are calculated to be affected by gerrymandering when compared to randomly drawn electoral boundaries.

The third problem is the vagueness of "Democratic voters’ clustering in urban districts." What does that mean? Given that House seats are allocated by population size (i.e. each seat represents roughly the same number of citizens), why does it matter whether Democratic voters are clustered or distributed as long as they are sufficient in number to win?

The only way I see that clustering in urban districts is relevant, and I don't think it is what Wasserman is getting at, is if by being clustered in urban areas, Democrats necessarily have materially different national policies than can be sold to the rest of the electorate outside those urban areas. For example, allocating Federal transportation taxes to rail commuting would be popular in urban areas (and therefor a Democrat policy) but would not be relevant or play well to all other voters not living in urban areas. On the other hand, a policy such as seeking to foster economic growth by reducing corporate taxes might be of interest to voters regardless of whether they lived in urban or suburban areas.

The point of all this is that by casting his analysis in a passive voice, Wasserman takes away Democratic agency and therefore clouds the root cause analysis. Occam's razor would suggest that the current Democratic woes can be sourced to their policy choices and not gerrymandering and clustering.

Wasserman continues,

But there’s a larger, long-term trend at work too — one that should alarm Democrats preoccupied with the future of Congress and the Supreme Court.

In the last few decades, Democrats have expanded their advantages in California and New York — states with huge urban centers that combined to give Clinton a 6 million vote edge, more than twice her national margin. But those two states elect only 4 percent of the Senate. Meanwhile, Republicans have made huge advances in small rural states — think Arkansas, North and South Dakota, Iowa, Louisiana, Montana and West Virginia — that wield disproportionate power in the upper chamber compared to their populations.

Well, yes. But you can't look at just one side of the equation (the Senate), you have to consider the House as well. What you give up in one, you might gain in the other. Our Constitution was born out of a series of compromises, one of them being a bicameral Congress with a Senate to ensure small states had a say in legislation and a House to ensure that there was a proportional voice for the populace.

New York and California are indeed at a disadvantage in the Senate with only 4% of the senators but they are a huge influence in the House with 27 and 53 seats respectively, i.e. 80 seats out of the total of 435 in the House (18%). Anyone wanting to pass legislation had better have California and New York onside. If the state caucus speaks with one voice. But they don't and that is a central problem for Democrats. 9 out of 27 house seats in New York are Republican and 14 of 53 in California.

The structure of the federal government is no late developing surprise. Talk of gerrymandering and clustering masks the core issue. Democrats need to have policies, strategy and leadership which appeal to more voters.

Wasserman goes into the history of the parties and their relative strength and presents a couple of interesting maps.

Consider: In 1980, there were 18 states where the presidential margin was at least 5 points more Democratic than the national result, 18 states where it was at least 5 points more Republican than the national result and 14 states in between. Hypothetically, over three successive election cycles, all either party needed to do to win a Senate majority was win all 36 of the seats in the friendly states plus at least 15 of the 28 swing-state seats.

Today, Republicans don’t even need to win any “swing states” to win a Senate majority: 52 seats are in states where the 2016 presidential margin was at least 5 percentage points more Republican than the national outcome. By contrast, there are just 28 seats in states where the margin was at least 5 points more Democratic, and only 20 seats in swing states.

What I think Wasserman misses, in his diversion of attention towards gerrymandering and clustering, is just how powerful Democrats were in 1980. And, really, for the sixty-two years from 1933 up until 1995. See Party Divisions of the United States Congress over time.

Between 1933 and 1995, 62 years and 31 Congressional sessions, Democrats were in the majority of the House 29 times or for 58 years. There were in control of both chambers 25 times, for 50 years. Gerrymandering, in that time, was substantially a Democratic Party activity.

So what changed? We won't know for a good while. Certainly the Republicans' Southern Strategy, the targeting of the South as it liberalized and modernized its economy, was a major factor in this realignment. Democratic Party neglect of its historical working class voters clearly has been a factor. Quite possibly the excess influence of celebrities, Wall Street and academia on Democratic Party policies has been a factor.

As I indicated above, Nate Silver and his FiveThirtyEight crew do good work and I believe with integrity of intent. This is actually an interesting and reasonably informative article. By taking the passive voice approach and leaning on old tropes like gerrymandering, however, I think Wasserman misses the larger canvas.

Yes, Democrats are in the political wilderness for now but things change quickly. Ronald Reagan's California was reliably Republican. Jimmy Carter's Georgia was reliably Democrat. Both concepts seeming alien and improbable if your context is only the past ten or twenty years.

Our governance depends on healthy competition between parties. Wasserman is right that the odd are stacked against the Democrats for the time being. I think what has happened is that the Democratic Party has temporarily come under the commanding influence of ideologues and academics such that their policies are driven by postmodern critical theory (identity politics, command and control economy and decision-making, repudiation of shared American cultural principles, etc.). My suspicion is that Democrats will remain in the wilderness for some time absent some surprising combination of exogenous events and/or Republican self-destruction.

But eventually they will decide that they desire power more than they do ideological purity and will adopt more American culture compatible policies and return as a political force. Republicans are such a big tent party that they will be vulnerable to a concerted push by a competing party with beneficial policies attractive the electorate at large.