the world. The social sciences have a long tradition of comparative research and can usefully apply the conceptual and methodological tools they have developed to the problem of global environmental change.
3. Human dimensions research should prominently include comparisons of human systems that vary in their environmental impact. Comparisons between countries or localities or of the same place at different time periods can show why some social systems produce as much human welfare as others with less adverse impact on the global environment. A number of important issues lend themselves to comparative and longitudinal approaches, including:
the causes of deforestation (studies can compare deforestation rates in countries that vary in their land tenure systems, development policies, and governmental structures);
the effects of imperfect markets on release of greenhouse gases and air pollutants (studies can compare the emissions of countries or industries with different regulatory or pricing regimes);
the sustainability of different agricultural management systems (studies might compare nearby localities in the same country);
the effects of different industrial development paths on fossil fuel demand (studies might compare time-series data for different countries);
the determinants of adoption of environmentally benign technologies or practices (for example, studies might compare industries or firms that do and do not recycle waste products);
the relationship of attitudes about environmental quality and materialism to environmental policies in different countries.
Such studies can ''unpack'' broad concepts, such as technological change, economic growth, and population growth, that are frequently offered as explanations of how human activities cause global change. Comparative studies offer the best way to get inside the broad concepts and identify more specifically the features of growth and change in human activity that drive environmental change.
4. Researchers should study the causes of major environmental changes both globally and at lower geographic levels. Global aggregate analysis may show a very different picture from analyses at lower levels of aggregation. It is important to have both pictures because aggregate data can obscure the variety of causal processes that can produce the same outcome. For example, the global relationship between economic growth and greenhouse gas