1. Data for needed variables included missing values
2. #1 filtered for missing values
3. #2 arrayed by year

Literature

Barrow (2000) discussion of Rule of Law, Democracy, and Economic Performance

Links

World Resources Earthtrends
Freedom House site
Freedom House data on free, partially free, not free
Heritage Foundation economic freedom reports

Comments on ACT variables

1. City economic intensity, GDP/km2: This is a measure for cities. For our national analysis, it may be more appropriate to use km2 for nation rather than for city.
2. Capital abundance, K/L: Use for nation as in ACT.
3. Lagged per capita income: this is GNP/population, not GDP/population. ACT used a the lagged 3 year moving average of the variable.
4. TI. Use the Openness variable in PWT 5.6
5. Interaction terms with TI: capital abundance, lagged per capita income, and interaction of ca and lpci.
6. See ACT data appendix listed on A’s website for Suburban Dummy, Rural dummy, CC dummies and interactions, Average temperature, Precipitation variation, and Helsinki protocol variables.
7. Following Barrow (2000, see tables 2.2 and 2.3 in particular), there may be better variables to include in the analysis than CC. Please read the analysis and commnet on the selection of the following as potenially importance variables in an analysis:
a. 5 year lag of electoral rights index (Freedom House)
b. Real per capita GDP (use GNP as income measure)
c. Years of primary schooling (primary, average length of schooling, (WRI Earthtrends)
d. Gap between male and female primary schooling (difference between c for males and females, WRI)
e. Ubanization rate (WRI)
f. Heavy reliance on oil production (source? WRI trade in energy exports?
(The argument for political freedom variables entering the analysis: ACT theory shows that political weights are important to efficiency of regulation. Literature suggests that political weights influenced by democracy and rule of law. Barrow shows that democracy and rule of law influenced by above variables. Hence, above variables should influence regulation and be important in determining pollution/water use. Of course, GNP is already recognized by ACT as an important variable)
8. Also, rather than Barrow’s variables in 7. We might create dummy variables for free, partially free, and not free from Freedom House data set. The latter are the summary data produced by FH. Where 1970 is not available, use data for 1972. Go to web-site and download the data: http://www.freedomhouse.org/ratings/index.htm
9. Also, as future variables
a. All 6 water and sanitation variables in WRI data.
b. GDP percents for industry, agriculture, manufacturing, services.
10. Create appropriate variables for humidity.
a. Relative humidity can be calculated from temperature and vapor pressure data (I believe). Look up in weather text or web for formulae. 10 year means for temperature and vapor pressure are available from http://ipcc-ddc.cru.uea.ac.uk/obs/index.html
b. Alternative: contact a agricultural climatologist and ask about sources of international relative humidity data