# Why does density predict political preference?

In The Righteous Mind, Jonathan Haidt cites Dave Wasserman,

Our counties and towns are becoming increasingly segregated into “lifestyle enclaves,” in which ways of voting, eating, working, and worshipping are increasingly aligned. If you find yourself in a Whole Foods store, there’s an 89 percent chance that the county surrounding you voted for Barack Obama. If you want to find Republicans, go to a county that contains a Cracker Barrel restaurant (62 percent of these counties went for McCain).

It turns out that there’s a decisive break between Democratic and Republican support at a population density of 800 persons/per square mile. Why might that be? Colin Gordon suggests,

There is, of course, a lot going on here, including a long history of regional and metropolitan patterns in partisan alignment. But a hopeful reading of the map would go something like this: People who live close to one another are more likely to know someone of a different color, a different income group, or a different sexual orientation. They therefore rely upon and appreciate the provision of public goods and public services (transit, parks, garbage collection), even as they consume fewer public dollars than their less-densely populated counterparts.

For the same reason that virtually every densely populated urban area in the United States overwhelmingly votes for Democrats: because Republican principles of governance would leave highly developed urban areas with insufficient tax revenue to maintain the infrastructure services without which those densely populated areas could not continue to support human habitation at those densities. (For those who don’t believe this, consider this: the third rail of Chicago city politics is snow removal, and an administration that fails at snow removal will not survive the next election.)

Population density is one of the best predictors of partisan voting outcomes: the higher the population density of a voting precinct, the more likely that precinct will vote Democratic. Indeed, the only large cities in the US that do not reliably elect Democrats are cities (such as Indianapolis, Indiana and Jacksonville, Florida) that have incorporated large tracts of lower-density suburban territory into their municipal boundaries. Manhattan is the most densely populated tract of land in the United States, and thus on that basis alone one can predict—correctly—that Manhattanites will reliably elect Democrats.

It’s a compelling argument, but it could be mistaken in one of two ways. First, greater density generally implies economies of scale, so government outlays per capita for infrastructure might be lower than in rural areas. For instance, Internet access is more costly in rural areas than urban areas.

Perhaps Martin’s unstated premise is that the output of critical infrastructure specific to dense urban areas is above the minimum long run average cost, and suffers from diseconomies of scale. Given Tiebout competition, cities might offer value in many other ways that attract and retain residents, despite an inability to minimize the long run average cost of infrastructure.

Second, even if urban-specific infrastructure would benefit from economies of scale, it’s possible that there are systematic ideological biases that obscure this knowledge from urban Democratic voters.

Might it be that the poverty and homelessness in plain view in dense cities embolden a politics of empathy via availability bias?

Might it be that cities attract and breed people who have high openness to new experiences? Linda Krueger on Quora supposes,

Those in urban areas are simply exposed to more people, more diversity, more cultures, more ideas. They know the world is made up of differing religions, beliefs, and lifesyles because they see it outside their door every day and coexisting peacefully as everyone moves through their day.

Rural areas are isolated from much of the world because the world and all its diversity is not right outside their doors. Instead they get to create a view of others and of opposing beliefs that will not be challenged to any great degree simple because their environment is more homogeneous. In that scenario fear is built on the unknown. They can assume immigrants are taking jobs even though none have taken their jobs. They can assume terrorist are just waiting to destroy them because they fear and are leery of outsiders. They can assume homosexuality is a curse to a moral society because a foundamental religious foundation gives them a sense of community, a strength of righteousness, and a guide to follow that comes from a high authority they can rely on and refer to.

Without any definitive answers, I’d guess that a difference in tolerance and openness to new experiences is what’s primarily driving the effect, and the other political beliefs follow via irrational clustering.

1 Based on research by David Wasserman of The Cook Political Report, reported by Stolberg 2011, as cited in Haidt, Jonathan (2012-03-13). The Righteous Mind: Why Good People Are Divided by Politics and Religion (Kindle Locations 6731-6732). Knopf Doubleday Publishing Group. Kindle Edition. Haidt, Jonathan (2012-03-13). The Righteous Mind: Why Good People Are Divided by Politics and Religion (Kindle Locations 5210-5213). Knopf Doubleday Publishing Group. Kindle Edition.

CC BY-ND 2.0-licensed photo “night density” by urban feel

# We really do need tolls on our roads

I was recently driving across the state of Virginia, and a trip that should have taken me about four hours via the highways actually took me about seven, due to heavy congestion. This is so beyond unacceptable for modern, civilized society. Such congestion is not inevitable.

Most road warriors I talk to about the problem of traffic seem to think the problem is intrinsic to our road system, which is simply not true. The existence of such horrible traffic frustrates me enormously, but each conversation with people who take it as a given compounds my frustration even more.

Just a little bit of basic economic theory points to an extremely obvious solution, but discussing such solutions triggers predictable political irrationality. A fee-for-service model is an obvious way to correct a commons problem, but for temperamental reasons, some individuals lack the ability to think systematically or partake in proper epistemic hygiene.

## Congestion is not inevitable

I can’t stress enough that traffic congestion is not inevitable. Some common refrains include: “There are just too many cars on the road,” or “This is just rush hour traffic.” Such fatalism is unwarranted.

Even meta-political public choice fatalism usually reserved for the federal government is unwarranted, because policy changes only need to be passed through state and local municipalities. Local governments are subject to Tiebout competition.

While unexpected accidents can always cause traffic jams, we can completely eradicate cyclical “rush hour” traffic.

## A commons problem

Congestion on highways is a clear instance of a tragedy of the commons. Without tolls, highways are open-access resources. Individual motorists using open-access highways aren’t paying for the full social cost of consuming the scarce resource at the point of service.

When a motorist takes a car onto an open-access public road, they enjoy the private benefit of using the road to travel to their destination, but they impose a public cost of increasing the probability of traffic congestion. The cost to any one individual motorist is low; no one car causes heavy congestion. When too many motorists add in their externalized costs onto the road, congestion results. It’s a giant prisoner’s dilemma, because no one motorist faces the incentives to limit their usage, and motorists have no way of cooperating with each other to ration out the resource in any logical way.

Tolls are a way of exacting the costs from each motorist adding one more car onto the road.

## The efficiency gains come about from changing behavior

Unsophisticated opponents of tolls have argued that cars unequipped with E-ZPass have to stop at tolls to pay, and actually cause buildups.

Where have we seen this flawed mental model before? Marx theorized that the base unidirectionally influenced the superstructure. Marx would have argued that tolls, as technologies in the base, constrain motorists by extracting money out of them. Marxist analysis can’t account for motorists’ individual cognitions. Dialectical materialism fails.

It’s the sort of mistake a child makes before developing a theory of mind. In reality, different motorists have different preferences and resources, and don’t want or need to use roads at the same quantity.

With tolls in place, individual motorists can decide on the margin whether or not to bear the cost of adding one more car to the road.

## Elasticity

In technical economic terms, a toll’s fare needs to be high enough to change the price elasticity of demand from being inelastic to elastic. Increases in fares need to rise so that the percentage change in quantity demanded is greater than the percentage change in the price. If the price of a toll isn’t high enough, then the price elasticity of demand would remain inelastic, and cyclical traffic would persist.

Critics of tolls charge that the poor bear the costs of tolls disproportionately. Such a claim is incoherent.

It’s important to remember that the problem the poor have is that they don’t have enough money. If we’re concerned in society with redistributing from rich to poor, we should redistribute the most liquid resource available, which is money, in the form of a transfer as close as possible to a lump-sum subsidy. Redistribution does not justify social engineering via taxes and subsidies on specific goods, services, or behaviors.

It doesn’t make sense to justify the continued existence of a market inefficiency on the basis that it puts the poor on an “equal” footing with the rich. At first glance, it may seem more fair to keep the resource unpriced. The underlying narrative is that not even the rich would be able to bypass the traffic in an open-access resource, and that if the rich were able to pay to exit, they would leave the poor either stuck in traffic, or unable to afford the new price of the tolls.

The narrative is wrong because when externalities exist, Pareto improvements are available by internalizing such externalities.

Tolls would actually help the poor, because the poor, along with everyone else, would experience the roads without traffic. Time wasted in traffic has the opportunity cost of foregone wages, commissions from sales, visits to clients, or valuable leisure time. Traffic also wastes gasoline. Dissolving traffic would increase the amount of resources available to all motorists, including the poor.

With the roads clear of congestion, people could still choose to substitute money for time by deciding how much to work. It makes no sense to lock people into a system that precludes any deliberate tradeoff, and it makes no sense to lock people into a system that just wastefully and unidirectionally converts money into time.

We can model the market for a toll road as having two segments, rich, R, and poor, P. The rich’s price elasticity of demand for a toll road ($E_{d_R}$) might remain inelastic, while the poor’s price elasticity of demand ($E_{d_P}$) might turn elastic:

• $-1
• $-\infty

This is a feature, not a bug. Income is one of the factors that affects elasticity. By converting the price elasticities of some motorists from inelastic to elastic, we can reveal preferences. Without any sort of bureaucratic central planning, motorists would find creative solutions to use toll roads at the point where marginal costs equal marginal benefits. Some motorists would turn to carpooling or slugging. Some would choose to not use the road during rush hour, and instead use the road only when the fares are lower. Some employers would adjust employees’ schedules accordingly.

## In the interest of fairness, don’t we need tolls installed everywhere, all at once?

Some critics charge that if tolls are only installed on a few key roads, the motorists who use the toll roads bear some cost, while the motorists of other open-access roads avoid such a cost. Such thinking is exactly backwards.

Money, for all its usefulness in fueling civilization, is still abstract enough to severely confuse people. The fact that money is exchanging hands doesn’t automatically mean that costs are higher. On the contrary, tolls internalize the externality of traffic, and thus, using a decongested toll road costs less than a clogged open-access road. A congested road has the high cost of wasted time; a decongested toll road has a lower cost that’s exactly reflected in the price of its fare.