There are certain things that are such staples that they defy the laws of price and demand.  The higher the price, the more we demand them.  This vicious cycle defies all seeming logic, but is the natural result when a good is so useful that people will choose more of it over other substitutes.  The substitutes just aren’t good enough, and since the staple is more expensive, we just need to buy more of it and less of them.

The famous example that comes to mind is potatoes in Ireland.  As potatoes became more expensive, the poor would sop end more of their food budget on potatoes, and less on vegetables, bread, cheese and meat.  Since they still needed enough food for a day, they bought more, not fewer, potatoes.  Potatoes were more still more marginally affordable, so they were more sensible than extravagant meat or value-added bread for the budget conscious poor Irish family in the 19th century.  Another model was starchy foods in the lives of the working poor.  Increases in price in these staples would ruin budgets for better foods like vegetables, eggs, or meat.

The question is, is traffic a Giffen good?  I looked at the relationship between Vehicle Miles Traveled (Consumption) and the annual cost of ownership, new car price and used car price (measures of cost) as a way to answer this question.

GiffenNewCar GiffenUsedCar 

For the data I have from the FHWA, there is a strong correlation indeed, especially before 2004.  After 2004, the correlations get weaker. Looks like a Giffen good!

Then again, there also appears to be a strong relationship between demand and price.  Looks like a standard market good, not a Giffen good at all!

Clearly, annual summary data is not sufficient to tease apart the causal link that distinguishes a Giffen Good from a Market Good.  But why not try?  I looked at the annual changes for cost and consumption, looking for the relationship between last years increase and next year’s.  Assuming people take a year to make decisions about consumption of traffic (which they of course don’t), this should allow me to hint at whether traffic is a Giffen or market good.


These are mostly weak correlations, although it does look like people are sensitive to new car prices.

Another condition of Giffen goods is that they be “inferior” goods.  That is, would people prefer to use other, more expensive things, if they could afford them.  It is hard to argue that traffic is inferior on convenience, range, speed, installed infrastructure, or privacy.  Traffic is demonstrably inferior on many other things, like energy, expense, danger, space, water and air pollution, real estate and social costs, however.

Is traffic a Giffen good?  No, it does not appear so, at least with annual data.