Hacker News Reader

Comment by yonran | Hacker News Reader

I think Emily Badger sort of misses the point. A scarcity tax is completely different from a user benefit fee for road maintenance. For public goods, it’s a good thing for the government to subsidize the fixed costs of the public good. But where there is traffic, the road is not entirely a public good anymore and we are already paying the cost of congestion, but the price is in the form of time and inconvenience rather than dollars. A congestion charge converts the time price into a monetary price. It makes sense to charge a congestion charge because it is merely harnessing a price that drivers already pay and using it to improve the city.

The same is true for land value taxes. Tenants pay the scarcity price for land anyway (in the form of land rent) regardless of the level of tax, so the government might as well harness that energy and use it to benefit all the people instead of a few.

View on HN

This assumes that either:

1. People can arbitrarily convert time into money, ie, there’s not a reason they’d prefer to pay some expenses in time.

2. People are free to change their travel times, and aren’t being coerced into traveling at a particular time. If travel time is a monopsony, then this is mere rent extraction because people can’t adjust behavior in response to economic pressures.

Yes, and in my opinion those are both valid assumptions to make.

1. You’re right, congestion pricing forces drivers to reveal the marginal price for time. In a congested road, what makes the marginal driver decide not to drive at a specific time? Because he knows that it will take too much time due to traffic (or because he is simply forced to queue). In a toll road, what makes the marginal driver decide not to drive at a given time? Because he knows it will cost too much money. At the margin, congestion pricing turns a time cost into a monetary cost.

For those individuals whose value of time is lower than the marginal driver (i.e., people who would rather have a traffic jam than pay the toll), congestion pricing itself can make them worse off, but when the proceeds are reinvested in transit I think it’s probable that congestion pricing will generally benefit lower-income people.

2. Congested roads already force people to change their travel time. If you try to cross a congested bridge at a specific time, you will be queued until all the people in front of you have passed. So I suppose you could say the road already is extracting rent and throwing it away. Congestion pricing harnesses the value (of all the drivers whose value of time is above the marginal driver) instead of discarding it.

I’m not sure I follow your argument on 2.

If people aren’t free to adjust their driving habits, congestion pricing merely raises the cost for them, because the time to commute doesn’t go down (appreciably) — everyone is still on the road at the same time.

It then effectively becomes a regressive tax: those least free to change their schedule are those in the lowest earning jobs, and hence those most coerced to pay the cost.

I’m not sure I believe that’s an effective way to raise money for government.

I disagree with your premise. There are people on the margin who decide to drive or not because of the traffic (or who are forced to queue unexpectedly). If the roads and parking garages had infinite capacity, then there would be hundreds of thousands more cars on the road in Manhattan. It’s implausible that all the people who remain on the road today are purely inelastic.

I also disagree that the tax would be incident on those with the lowest paying jobs. Higher-paying workers are more likely to pay more, since people with price-inelastic demand for driving are more likely to be those who value their time more highly.

Finally, I don’t think the primary purpose is to raise money for the government, but instead to manage the road resource so that the throughput is high.

That decreasing the price causes elastic demand doesn’t indicate that raise the price decreases demand elastically: there can be a largely inelastic base load.

> Higher-paying workers are more likely to pay more, since people with price-inelastic demand for driving are more likely to be those who value their time more highly.

Again, this is only true absent coercive forces —

Higher paid workers (on average) also have more schedule flexibility and ability to use money to cover for their physical absence.

By contrast, lower paid workers have less flexible schedules and a greater need to utilize their time — eg, they have to perform childcare because their market rate per hour is below child care services.

This leads to a situation where high paid workers have more elastic price demand than low paid workers, and raising prices ends up charging low paid workers because if market distortions from other factors.

Reply on news.ycombinator.com