Sunday, August 14, 2005

Do Greens Buy Hummers? Free Riders versus “Good” Environmental Citizens

Economic theory predicts that common property is trampled upon and seized. Public parks, the Atmosphere, Oceans, public libraries all are polluted in part due to the fact that nobody has an explicit monetary incentive to preserve them. Each new semester, we teach students about the “Tragedy of the Commons” but does everyone free ride?

This summer I saw the net results of free riding in Rome. I spent all of June in Rome and was shocked by both the rampant dog poop on the streets and the graffiti covering so many buildings. I wondered why “quality of life” laws are not enforced in this city (maybe Rudy Giuliani should serve as Rome’s mayor?). What percent of Rome’s citizens engage in defacing the Commons? Turning this question around; in the United States, who buys hybrid cars such as the Prius and the Insight and what motivates them to do so? It is well known that relative to conventional vehicle, hybrids create less smog and less greenhouse gases. But what motivates buyers? Intrinsic environmentalism? High gas prices? Or, the Warm Glow of being seen doing a good deed?

A central theme in modern applied micro economics is recognizing that people differ. Labor economists and behavioral economists are hard at work documenting population heterogeneity with respect to earnings, ability, effort, patience, and other dimensions. About a year ago, I started working on an “environmental heterogeneity” project. In my empirical study (Environmental Ideology as a Determinant of Household Resource Consumption), I wanted to identify people who care deeply about the environment (Greens) and compare their market choices to the market choices of people of similar income and age but happen to not care about the environment (the Browns). Intuitively, I wanted to compare Paul Ehrlich’s daily choices to Dick Cheney’s daily choices. Relative to the Vice President, does Paul consume fewer natural resources in day to day life?

When I told people about this project, they told me that I was in a no-lose situation. If I found that Greens have the same private consumption patterns as Browns, then I could facetiously title the paper “Hypocrites”. If I found that Greens had a smaller per-capita ecological footprint than Browns, then some environmental groups would celebrate the paper saying that it represents objective evidence supporting the environmental movement’s success in educating people. This argument would claim that when people become “environmentally enlightened”, they choose to live their lives in a more sustainable way.

As I pursued this project, I learned that talented environmental economists had already been working on this issue using very different data than what I describe below. Matt Kotchen and Michael Moore (2004) surveyed Michigan households concerning their propensity to participate in a green-electricity program. They find that conservationists consume almost 10 percent less conventional electricity than non-conservationists and that conservationists are more likely to participate in the green-electricity program. How did they know who was a conservationist? By surveying each household, they asked questions well chosen to partition people into three sets (Greens, Browns, neither).

I approached this problem using a different data set. Unfortunately, very few household level data sets contain data both on how much resources different households consume and information about each household’s ideology. Intuitively, how do we know an “environmentalist household” when we see one? Some surveys ask people; “Are you a Republican?” This might be correlated with environmental ideology but maybe not.

To make progress on my core question, I used data solely from California. The 2001 National Household Transportation Survey provides detailed information for thousands of households on how much gasoline they consumed in the previous year. The data set provides demographic information about the household and provides the household’s residential zip code. Knowing each survey household’s zip code, I merged to each record the local community’s share of registered Green Party voters. The Green Party is a minor political party in California. In the year 2000, only .9% of the state’s voters were registered party members. At the state level, this party does not exercise political clout. Thus, it is difficult to view Green Party membership as anything but an expression of one’s core beliefs. I’m claiming that communities where a relatively large percentage of residents are Green Party members are pro-environment communities. For example during summer times, I live in Berkeley California in zip code 94707. The Green Party % registered for this zip code is 4.3%! Democrats have a 69% share here. To learn more about this Party take a look at:

A critic could claim that I only have a “noisy” measure of each person’s own ideological ideology because I only observe a measure of the average ideology in his local community. I agree with this point but it beats nothing! The proof is in the empirics.

The first part of the paper’s empirical findings documents that “Green Party” communities vote pro-environment on California Ballot Initiatives such as Prop 185 in 1994. Prop 185 would have raised gasoline taxes and spent the revenue on improving public transit.

To give a sense of some of my paper’s findings I’ll focus on some zip code level regressions I estimated by merging year 2000 Census of Population and Housing data with the zip code data on political party registration. The Census collects information on commuting modes and I used as the dependent variable the percent of commuters who walk or bike. Controlling for zip code average household income and population density, Democrats are no more likely to walk or bike than Republicans while Green Party communities are more likely to commute by walking or biking. In the entire state of California, only 3.9 percent of workers commute using this “green” technology. A two percentage point increase in an area’s share of Green Party registered voters increases the share of commuters who walk or bike by 3.4 percentage points. Again, this is a large effect. The right column of Table Four switches the dependent variable to the share of a zip code’s workers who commute by public transit. Public transit is used more in high population density, poorer zip codes. Controlling for these variables, both Green Party communities and Democrat Party communities are more likely to commute by public transit than Republican communities. Note that the Green Party coefficient is more than 10 times larger than the Democratic Party coefficient.”

For folks who are interested, I will post a .pdf of the entire paper to my Fletcher Web page in September. The bottom line of my study is that Greens vote green and consume green. Areas filled with Democrat registered voters vote green but have the same consumption patterns as Republican areas.

A weakness of my study is that there is no price variation. I cannot answer the question; “if gas prices increase by 20% will Greens reduce their consumption of gas by 42% while Browns will reduce their consumption by only 12%”. Ideally, future research will come up with ways to estimate how demand curves for green products differ across the population. As for profit firms such as Toyota and Honda design “green cars” such as the Prius and the Insight, their marketing departments must be trying to figure out what prices to charge for these cars and what would demand be at those prices. At what price of gasoline and at what price charged by Toyota would Dick Cheney buy a Prius? At what price of gasoline would Arnold substitute away from his Hummer collection?