NBER papers on transport and labour economics

The NBER publish some good economics! Here are a couple of new working papers I liked the sound of :

Pounds that Kill: The External Costs of Vehicle Weight
Michael Anderson, Maximilian Auffhammer

Heavier vehicles are safer for their own occupants but more hazardous for the occupants of other vehicles. In this paper we estimate the increased probability of fatalities from being hit by a heavier vehicle in a collision. We show that, controlling for own-vehicle weight, being hit by a vehicle that is 1,000 pounds heavier results in a 47% increase in the baseline fatality probability. Estimation results further suggest that the fatality risk is even higher if the striking vehicle is a light truck (SUV, pickup truck, or minivan). We calculate that the value of the external risk generated by the gain in fleet weight since 1989 is approximately 27 cents per gallon of gasoline. We further calculate that the total fatality externality is roughly equivalent to a gas tax of $1.08 per gallon. We consider two policy options for internalizing this external cost: a gas tax and an optimal weight varying mileage tax. Comparing these options, we find that the cost is similar for most vehicles.

(See also my post on external accident risk)

The Incidence of Local Labor Demand Shocks
Matthew J. Notowidigdo

Low-skill workers are comparatively immobile: when labor demand slumps in a city, low-skill workers are disproportionately likely to remain to face declining wages and employment. This paper estimates the extent to which (falling) housing prices and (rising) social transfers can account for this fact using a spatial equilibrium model. Nonlinear reduced form estimates of the model using U.S. Census data document that positive labor demand shocks increase population more than negative shocks reduce population, this asymmetry is larger for low-skill workers, and such an asymmetry is absent for wages, housing values, and rental prices. GMM estimates of the full model suggest that the comparative immobility of low-skill workers is not due to higher mobility costs per se, but rather a lower incidence of adverse labor demand shocks.

In other words, policies that protect against unemployment and poverty also reduce mobility, which obviously doesn't mean the policies aren't worthwhile but does tend to increase the likelihood of both spatially concentrated and temporally extended unemployment.

UK house price to rent ratios aren't all that reliable

The Economist has a nifty interactive resource showing trends in house prices and associated indicators for a bunch of countries. It includes trends in the ratio of house prices to rents, which in theory should be useful as an indicator of over- or under-valuation in house prices - if prices are very high in relation to rents then that's a sign that they are over-valued as an asset in comparison to the income stream they can be expected to earn. According to the Economist's data, the price to rent ratio in Britain is still way above its historical average, and this is evidence that housing is over-valued as an asset.

That could well be true, but I think we should be cautious in how we interpret these figures, because constructing them is not at all straightforward, perhaps particularly in Britain. That's because while we have a bunch of house price indices we have no equivalent for market rents: that is, no reliable and comprehensive data on trends in average market rents, from either official or private sector sources. So what is the Economist using? They say the data is from Office for National Statistics but are no more specific, but since ONS to my knowledge don't have any dedicated market rent data my guess is that they are relying on the 'Rent' component of the official CPI/RPI indices.

The problem with that source is that it includes not just private sector but also social rented housing, i.e. council and housing association stock (see here for the 'basket' of goods and services included in CPI/RPI). But social housing rents are strictly controlled by government regulation, so the trend in social rents is typically quite different from the trend in market rents, making the CPI/RPI rental index a poor proxy for either.

So I would be much more cautious than the Economist, who confidently assert here that British house prices are 30% over-valued and do the same calculation for a bunch of other countries, the data for which may or may not be any better than for Britain. The benchmark used to identify over- or under-valuation is a long-run average, but even leaving aside the very large problem of data quality it is not clear to me that the level of the price/rents ratio in, say, the late 1970s is a good guide to what it should be in the 2010s.

Anyway, there you have it. Note that I'm not saying British houses aren't over-valued, just that you can't really tell from the market rents data we have at the moment. I have heard that the Valuation Office Agency will start publishing statistics based on their very extensive market rents database soon, which would be great, but that still won't provide the kind of historic trend the Economist's table requires.