Is Everyone Gaming the System?
Following on the heels of the news about Volkswagen and Mitsubishi noted in Octobers Editors Letter, we received a news release from the Toulouse School of Economics suggesting that the whole industry has rigged fuel economy tests in Europe to comply with strict carbon regulations. TSE Professor Mathias Reynaert and his co-author James Sallee from the University of California Berkeley analyzed data from Dutch company TravelCard NV that they claim show clear signs of test result fiddling industry-wide.
The authors measured the real fuel consumption of vehicles on the road and compared it to car manufacturers claims. As Reynaert explained, We observe from the data that real carbon dioxide emissions are much higher than in tests. In 2004, the difference between tested and real consumption was about 10 percent, which is normal. However, this gap has increased to about 40 percent over the last five years.
The data does not simply point to Volkswagen and Renault, but to all car manufacturers present in the data. According to Reynaert, this increase is clearly correlated to regulatory pressure that pushed firms to exploit loopholes in the test process: In 2007, as the EU announced that the emissions of new car models would have to decrease to 130 grams of carbon dioxide per kilometer, manufacturers had to adapt to this daunting target. However, the procedure used to establish official emission ratings, the New European Driving Cycle, leaves firms a lot of leeway to game the test results.
Reynaert believes the solution lies with legislators and regulators. To put an end to the gaming, the authorities need to introduce more realistic, randomized testing enforced by third parties.
In their just-published paper, Reynaert and Sallee use the empirical example of car emissions testing and a theoretical model centered on Goodharts Law to show that policies tend to break down the self-regulation ability of an industry. Goodharts Law is named after economist Charles Goodhart, and its original formulation was As soon as the government attempts to regulate any particular set of financial assets, these become unreliable as indicators of economic trends. This is because investors try to anticipate the effect of the regulation and invest so as to benefit from it.
The law is implicit in the economic idea of rational expectations. While it originated in the context of market responses, the law has implications for the selection of high-level targets in organizations. The researchers believe their model could prove relevant for other regulation issues. Read the full paper at nature.berkeley.edu/~sallee/reynaertsallee_gamingnber.pdf.