Research Team *
Thomas Dietz
Eugene A. Rosa
Richard York

*Listed Alphabetically

STIRPAT is an acronym that refers to a statistical model for assessing environmental impacts at virtually any scale and to a research program in structural human ecology (SHE). The STIRPAT model was recently utilized by the research team in an article titled "Driving the Ecological Footprint" (Frontiers in Ecology and The Environment 2007; 5(1): 13–18). This comparative analysis shows that population size and affluence are the principal drivers of anthropogenic environmental stressors, while other widely postulated drivers (e.g. urbanization, economic structure, age distribution) have little effect... More | Go to Full Article

Richard York wins the 2007 Outstanding Publication Award of the Environment and Technology Section of the American Sociological Association... More  

STIRPAT research team member Eugene A. Rosa, professor in the Washington State University Department of Sociology, was selected to give the Distinguished Faculty Address for 2007, the longest-standing award of distinction at WSU. Go to Full Article

The STIRPAT program of research was honored with the 2004 Outstanding Publication Award for distinguished scholarship in environmental sociology by the American Sociological Association Section on Environment and Technology.
 
Introduction: The Dynamics of Human Ecosystems

The human species is unique in many ways. One uniqueness resides in the complexity of human ecosystems and in the extraordinary role humans play in determining the sustainability of those systems. Hence, a central problem for human ecology is to understand how human actions influence environmental change. Structural human ecology seeks to understand the underlying structural drivers of environmental change and their dynamics.

Any human ecosystem comprises four key, interacting components: population, environment, social organization, and technology. How might we better understand the linkages between these indispensable components and environmental impacts? How might we proceed to develop organized research programs to examine these linkages? How might we discipline our conceptual models with empirical tests?

IPAT
One of the first empirical and quantitative efforts to address these questions was the accounting equation:

Impact= Population x Affluence x Technology, or I=PAT

This mathematical identity, referred to by the acronym IPAT, was developed within the framework of traditional ecology. That development was launched by a debate between Barry Commoner and his colleagues on one side and Paul Ehrlich and John Holdren on the other in the early 1970s. Over the following decades, the I=PAT equation was occasionally used to estimate the effects of human population, level of affluence and choice of technology on environmental impacts or to project future environmental change based on changes in the three driving forces. Recently IPAT was renamed the Kaya equation by the Intergovernmental Panel on Climate Change (IPCC). The Kaya equation has been very influential in analyses of energy consumption and greenhouse gas emissions. For further information on IPAT see the bibliography.
STIRPAT
Dietz and Rosa wished to capitalize on the analytic potential of IPAT in order to address the questions posed above. They recognized, however, that human systems, while organized by the same causal processes governing natural systems, were fraught with random influences and noise. Human activity seldom follows precisely mathematical functions that characterize non-human systems. Hence, In 1994 Dietz and Rosa proposed that the I=PAT accounting equation would be more useful if recast into a stochastic form that would allow random errors in the estimation of parameters and permit systematic hypothesis testing. They proposed the following reformulation of IPAT:
I=aPbAcTde
where I is a measure of environmental impact, P is population, A is affluence and T is technology. Then a, b, c, d and e are parameters to be estimated. Estimation is accomplished by using standard statistical methods, such as regression analysis. The regression estimation procedure requires the key variables—I, P, A, and T—be converted to logarithmic form. The resulting estimates of the parameters b, c, and d can then be interpreted as the elasticities of population, affluence and technology, respectively. The elasticities tell us the percent change in impacts from a one percent change in any of these driving variables. The “a” term is a parameter that scales the model. The “e” term captures the affects of all things that effect I but are uncorrelated with P, A and T.
Dietz and Rosa named the reformulated version of IPAT as STIRPAT, an acronym for STochastic Impacts by Regression on Population, Affluence, and Technology. The acronym STIRPAT was chosen on the basis of several considerations. First, we needed a label that distinguished the stochastic version of IPAT from the accounting version of IPAT. At the same time we wished to recognize and give proper credit to the origins of the basic idea that impacts are a function of population, affluence, and technology. We chose STIRPAT because we had “stirred” the original idea, had retained the original terms I,P,A,and T in deference to the founders of the original formulation, and because the word "STIRP" means a line of descent from a common ancestor (The Random House Dictionary, 1987).
STIRPAT is a research program in structural ecology. It focuses on macro drivers and macro impacts, such as CO2 loads and other greenhouse gases and loss of biodiversity with the nation state as the unit of analysis, Richard York joined the research team in 2000 by developing a line of research that focuses on the ecological footprint (the most comprehensive measure of impacts currently available) and that systematically tests a variety of economic and sociological theories predicting the trends in impacts.
Key Elements of the STIRPAT Research Program
-- Understanding the anthropogenic driving forces of environmental change is one of the fundamental scientific and practical questions of the 21st century. In the last decades of the 20th century, it became clear that humans were having a massive influence not only on most local and regional ecosystems but also on major biogeochemical processes, including climate. The pace, scale, and spread (PaSS) of human impacts in contemporary times is unprecedented. Nobel Laureate Paul Crutzen1 and ecologist Eugene Stoermer (2000), recognizing this unprecedented change in the global environment, have redefined the term “Anthropocene” to describe the current period that begins roughly at the time James Watt perfected the steam engine in the latter part of the 18th century2. The Crutzen/Stoermer refinement emphasizes the post-industrial impacts to the environment by the human species and locates the “Anthropocene” in the most recent stages of the Holocene era (the post-glacial geological epoch of the past ten to twelve thousand years). This telescoping of the modern era, this alignment with the age of modernity, underscores the “astounding” “anthropogenic” exploitation of the Earth's resources in the past three centuries.
--Understanding coupled human/ natural systems requires interdisciplinary analysis. Analyses that draw only on a single discipline or meta-disciplines (social sciences, physical sciences, biological sciences, engineering) may provide some insights into the structural dynamics of human ecosystems, but these analyses are at best partial and at worst mis-specified. The STIRPAT research program reflects this commitment to interdisciplinary analyses.
--There is value in work at all scales from the micro-level of individuals and households to the macro-scale of nation-states and the global system. Most work on driving forces is at the micro-level. Sociology has long recognized that the activities and dynamics of social systems—societies, institutions, cultural patterns— often cannot be reduced to the decisions and behaviors of the individual actors constituting those systems. Hence there has been a long tradition of macro-comparative work in the social sciences that has produced useful knowledge and frameworks for understanding the structural dynamics of human systems. Structural human ecology combines that tradition with a focus on anthropogenic environmental change.
--“All models are wrong. Some models are useful.”—George E.P. Box, (distinguished statistician). Models are crude approximations to reality and, as pointed out by Box, must be wrong in describing reality. An effective model is an abstraction of the real world that captures sufficient detail to be realistic but omits inessential detail that would complicate the abstraction needlessly. Hence, Models can be useful and are often an indispensable aid to research. But there are inherent limitations in the multiple uses to which models can be put. Nearly four decades ago the distinguished Harvard ecologist Richard Levins summarized these limitations: It is impossible to simultaneously maximize generality, realism, and precision. This leaves three options: (1) sacrifice generality to realism and precision; (2) sacrifice realism to generality and precision; or (3) sacrifice precision to realism and generality.
The STIRPAT research program begins with the simple IPAT model and expands from it to more complex and hopefully more realistic models. This aligns with a long tradition in science of using simple models as a starting point to structure analysis but being open to adding complexity. In sum, STIRPAT sacrifices some realism to generality and precision.
--Functional form matters. IPAT suggests a multiplicative decomposition of impacts. STIRPAT is a straightforward extension of that idea. In some social science analyses it is common to see a mixture of linear, polynomial and logarithmic terms introduced to transform variables toward normality. We are cautious about producing functional forms that are not interpretable, as often happens with admixtures of transformations that are the result of a preoccupation with model fit rather than model insights.
STIRPAT Bibliography
Many of the analyses contained in the STIRPAT publications use the Ecological Footprint (EF) as a measure of overall human impact on the environment. The EF is the most comprehensive and most rigorously reviewed measure of overall impact of which we are aware. For more information on EF, please see The Global Footprint Network at www.footprintnetwork.org.
IPAT Bibliography
STIRPAT on Earth and Sky Radio: Interviews with Gene Rosa
http://www.earthsky.org/?q=Gene+Rosa
 
STIRPAT Research Team

Thomas Dietz, Environmental Science and Policy Program, Michigan State University
Bio
www.dietzkalof.org

Eugene A. Rosa, Department of Sociology, Washington State University
Bio
http://cooley.libarts.wsu.edu/rosa/

Richard York, Department of Sociology, University of Oregon
Bio
http://sociology.uoregon.edu/faculty/york.php

Related Links
1Crutzen shared the 1995 Nobel Prize in Chemistry with Mario Molina and Sherwood Rowland for basic discoveries of the effects of chlorofluorocarbons (CFCs) on the earth's ozone layer.
2The entire Holocene (Recent Whole) era, comprising the last 12,000 years, is sometimes labeled as the Anthropocene (meaning recent Homo Sapien Sapiens) to reflect, somewhat incorrectly, the emergence, survival, and dispersal of humans around the globe. Actually humans had evolved and dispersed all over the world prior to 12,000 years ago. Furthermore, ecological disasters up to three centuries ago tended to be isolated and localized, not a threat across the globe.