The data is in both plain text (TXT) and Lotus 123 (WK1) format. There are three files. In ENV_Y1 the data is sorted by the (World Bank's) country code and year. In ENV_Y2 the data is sorted by country code, city, pollution measurement site type, and year. In ENV_Y3 the data is organized by country code, river, city, and year. The list of variables in the three files is followed by detailed description of the variables used in the paper. Also see Annex A of the paper.
Variables used in the paper are in natural logs unless otherwise noted below. The variable names followed by file names in parenthesis refer to raw data. Where necessary the derivation formula for the derived variables are included in the description. The raw data includes all outlier data.
Publication date: October 10, 1992
List of variables:
- Lack of safe Water (watti/ENV_Y1) was measured as the percentage of population without access to safe drinking water. In urban areas access to safe water was defined as access to piped water or a public standpipe within 200 meters of a housing unit. In rural areas, it implies a family member need not spend a disproportionate part of the day fetching water. "Safe" drinking water includes untreated water from protected springs, boreholes and sanitary wells, as well as treated surface water. Data for this measure was available for only two years, 1975 and 1985 for 44 and 43 countries respectively.
- Lack of urban sanitation (sanui/ENV_Y1) was measured as percentage of urban population without access to sanitation. Access to sanitation was defined as urban areas served by connections to public sewers or household systems such as pit privies, pour flush latrine, septic tanks, communal toilets and other such facilities. Data was available for 1980 and 1985 for 55 and 70 countries respectively.
- Annual deforestation (forest/ENV_Y1) reflected the yearly change in forest area for 66 countries between 1962 and 1986.
- Total deforestation (forest/ENV_Y1) the change in forest area between the earliest date for which substantial data was available, 1961, and the latest date, 1986. Total deforestation data was available for 77 countries.
- Municipal solid waste per capita (mwpc/ENV_Y1) computed in kilograms, on the basis of available city level information for 39 countries compiled for the year 1985.
- Carbon emissions per capita (co2pc/ENV_Y1) expressed in metric tons per person per year, for 118 153 countries between 1960 and 1989.
- Sulfur dioxide (so2/ENV_Y2) Data on ambient levels of sulfur dioxide measured in micrograms per cubic meter were available for 47 cities distributed in 31 countries for the years 1972 to 1988.
- Suspended particulate matter (spm/ENV_Y2) Data on ambient levels of suspended particulate matter measured in micrograms per cubic meter were available for 48 cites in 31 countries for 1972 to 1988.
- Dissolved oxygen (dox/ENV_Y3) measured in milligrams per cubic meter, was available for 57 rivers distributed in 27 countries for intermittent years between 1979 and 1988. Dissolved oxygen measures the extent to which aquatic life can be supported. Low levels of dissolved oxygen can result from large amounts of industrial effluent or fertilizer runoff from adjacent agricultural land.
- Fecal coliform (fcol/ENV_Y3) measured in numbers per 100 milliliter, was available for 52 rivers distributed in 25 countries for intermittent years between 1979 and 1988. Fecal coliform measures the level of biological refuse in the river water. High levels of fecal coliform are associated with high incidence of water borne disease in the affected area. Data from five rivers were excluded from the sample due to extremely high reported levels of fecal coliform (exceeding 700,000 per 100 milliliter). These rivers are Atoyac, Balsas and Lerma in Mexico, San Pedro in Ecuador and Yodo in Japan. The effective sample size for fecal coliform was reduced from 434 to 402.
- Income per capita (gdpch/ENV_Y1) Real per capita gross domestic product in purchasing power panty (PPP) terms were used for the years 1960 to 1988 for 95 138 countries (variable RGDPCH in Penn. World Table Mark 5.). The chain base method of indexing was used to take into account the changing production bundle over the period. GDP data was not available for all the countries for all the years.
- Growth of real income (gdpch/ENV_Y1) derived as the log of the first difference.
- Investment as share of GDP (i/ENV_Y1) Investment data was from Summers and Heston, 1991 in PPP terms (variable "I" in Penn. World Tables Mark 5).
- Electricity Tariff (etrf/ENV_Y1) in US cents per kilowatt hour for 60 countries in 1987 was used as a proxy for energy prices in those countries between 1985 to 1988.
- Trade share in GDP (tvalx/ENV_Y1) Trade share was measured by total imports and exports of goods and non factor services as a percentage of GDP. Data were available for 67 88 countries between 1960 and 1988.
- Parallel market foreign exchange premia (black/ENV_Y1) based on the difference between official exchange rates and parallel market rates. The data was available for 36 99 countries for the years 1960 to 1988.
- Dollar's outward orientation index (dollar/ENV_Y1) for 87 90 countries between 1973 and 1985 as derived in Dollar, 1991. The index was computed by the weighted average of mean price distortion in the period 1973 85 and of its standard deviation. The price distortion was calculated as the residual of a regression of the relative price of consumption goods on urbanization, GDP per capita (a proxy for a country's endowment) and an interactive term. All variables were entered in logs, so that the residual could be interpreted as a percentage of deviation from an appropriate level as determined by the countries' endowment.
- Debt as a share of GDP (dbt_y/ENV_Y1) for 83 119 developing countries between 1960 and 1988 was used. Debt was defined as disbursed amounts of short and bag term external liabilities outstanding and IMF credit. GDP was measured in 1987 us dollars.
- Political rights index (gas_pol/ENV_Y1) measures rights to participate meaningfully in the political process for 108 119 countries for 1973 and "1975 86 on a scale of one to seven where lower numbers indicate greater political rights (see Gastil, 1989). A high ranking country must have a fully operating electoral procedure, usually including a significant opposition vote. It is likely to have had a recent change of government from one party to another, an absence of foreign domination, decentralized political power and a consensus that allows all segments of the population some power. The index was constructed on the basis of satisfaction of the above and other related criteria by the countries in question. This variable is an index and not in logs.
- Civil liberties index (gas_civ/ENV_Y1) measures the extent to which people are able to express their opinion openly without fears of reprisals and are protected in doing so by an independent judiciary. Though this index reflects rights to organize and demonstrate as well as freedom of religion, education, travel and other personal rights; more weight was given to those liberties that are most directly related to the expression of political rights. This variable is an index and not in logs.