Three Different Approaches to Community Health

At present, one can organize the community health science literature into three dominant approaches.

1. Social Pathways Model: The oldest and most widely practiced approach is the social pathways model. This model takes a nomothetic position, seeking to determine how a small set of social factors impacts the health of a community. In this model, community is also treated as a dependent (or grouping) variable.

2. Community as Context Model: This more recent approach emerged during the 1990s and has remained very hot! In this model, community context is treated as an independent variable, separate from the contribution of various other social factors--income, educational level, family health behaviors, etc. This approach to studying communities is a top-down model.

3. Community as a Complex System: The last model is the newest and least practiced. It views communities as complex systems; and takes a bottom-up approach to modeling.

The strength of the third approach is its ability to overcome the limitations of the other two models.

The other two models suffer from a reductionistic approach to community health--community is either an independent or dependent variable, with little research done to explore the "system-level" effects of a community; or, for that matter, the link within a community between micro-level (agent-based) and macro-level (emergent) behaviors. There is also no sense of environmental forces or the dynamics of a community over time--as a system--in the other two models.

Obviously, the limitations of the first two models are challenges that a complexity science approach to communities can handle. It can handle these challenges because this third approach has a complex view of communities as systems--that is, it sees the link between the micro and macro; has the tools to study system-level, emergent behavior; and has the ability to frame how environmental forces and the larger systems within which communities are situated impacts their respective health. Its bottom-up approach also allows it to see communities as both independent and dependent variables (via the concept of feedback loop). And, its bottom-up approach allows it to see communities as both context and composite--in other words, it does not construct a false dichotomy between community and other social (individual-level) factors such as income, education, etc.

For a basic introduction to the community-as-complex-system model, see Tim Blackman's new book, Placing Health.


  1. Very interesting! If I am understanding the idea of community-as-complex-system correctly, by extension, you could take this to the next level and treat all the communities in a county as agents in a larger complex system in order to look at the dynamic characteristics of the county as a whole.

  2. Thanks for the post. Actually, following your argument, you could do one of two things. You could merge a particular community into a larger system--say, for example a county. This way, the community would become an agent in the larger system. For example, 10 communities within a larger county; or, by extension, 50 states within a country.

    The other thing you could do is keep you analysis at the level of the community and then position it within, as you say, larger systems. So, you could study a particular community within a city, then this city within a county, and then this county within a state, within a country, etc.

    This later type of analysis is not for showing off. It is a legitimate necessity. We need more research that moves between the micro and the macro to understand how globalization, for example, effects the local level and, conversly, how the local impacts the global, and the various levels in-between.

    This, in fact, is why we created the SACS Toolkit--our new method for modeling social systems, which we discuss in our new book on sociology and complexity science. We knew that to analyze systems within systems within systems requires a strong organizational framework so that the researcher can manage all of that data.

    So, this is a longwinded way of saying, I think you make an important point.