Home Page (MVA) > Integrated Assessment Models (IAMs) and Resources > IAMs Thematic Guide
Thematic Guide to Integrated Assessment Modeling
[HOME] [PREVIOUS] [NEXT] [BOTTOM]
Introduction
Over the past few years Integrated Assessment (IA) has emerged as a new unifying approach for making progress towards a better understanding of problems in global change. Assessments of policies to respond to global climate change are increasingly being conducted using Integrated Assessment Models (IAMs). These models attempt to integrate information by linking mathematical representations of different components of natural and social systems in a computer model.
As integrated assessments become more accepted as legitimate approaches to providing policy relevant insights, we in the IA community can no longer treat our work as merely exploratory. If societies, and indeed the world, choose to base decisions at least partly on our recommendations then we have an enormous responsibility; for we all know how little we know and how little we can know[FN]. This is clearly a major challenge. On the one hand there is political (and motivational) pressure to provide policy guidelines. On the other, climate change IA is a nascent field, still striving to define itself and without widely accepted assumptions, norms or procedures.
To be sure, the lack of commonly accepted rules and procedures is not unique to IA. Many new research areas face similar quandaries. What is special about IA of climate change is its truly interdisciplinary nature, and the multiple spatial and temporal scales of relevant phenomena. Additionally, the promise of providing input to policy recommendations is seductive. Unlike many other areas of enquiry, the IA community cannot stop at merely doing analysis. We all have the hope that the culmination of our collective efforts will inform policy recommendations. As a necessary condition for informing policy, IA needs to satisfy certain robust criteria of internal quality control and appropriate use. As a community we need to establish norms and procedures that distinguish good analyses from bad ones, to be more reflexive about our own analyses, and to make all efforts to guard against our analyses falling prey to political expediency. Otherwise, we run the risk of not being heard at all, or of speaking with the wrong voice in the political clamor over climate change.
Many members of the IA community are acutely aware of quality issues. Morgan(1995) has laid out seven attributes of a good integrated assessment which provide a broad basis for performing integrated assessments. However, we believe that overall there has been little discussion in the IA community regarding quality control and what we as a community can do to address this important issue. This paper argues for the need to focus attention explicitly on quality control issues, and highlights some of the issues that arise in attempting to maintain quality in IA. We believe that IA is at a crucial period in its development, having emerged from relative obscurity into the limelight. This development accentuates the need for serious critical review and reflection.
In this paper we make the assumption that IA models have the potential to provide meaningful input to global policies with regard to climate change. We will not explicitly address the underlying basis for this assumption here. Further, we do not address issues relating to the communication and incorporation of knowledge from IAs in the policy process. Given the vast uncertainties and long time scales of the climate problem it could well be the case that IA efforts will not provide any concrete policy advice; their utility may be limited to educational and research purposes. In other words, the suggestions regarding quality control in this paper are intended to facilitate better IA studies, but there is no guarantee that even the most rigorous IA studies will produce knowledge that drives policy.
The remainder of this paper is organized in the following way. In section 2 we raise three important questions regarding the purpose and practice of IA. In section 3 we discuss experiences with other unifying approaches where expectations were belied by actual results. We examine the implications these lessons from history have for integrated assessment. In section 4 we examine some examples of integrated assessment, and use them to identify a number of problems with IA. These examples serve to further highlight some of the pitfalls of performing IA, and are not intended as indictments of particular analyses. In section 5 we lay out what we think are major causes for the pitfalls that IA faces. In section 6 we highlight a number of issues that we think the IA community needs to address before the full potential of IA can be realized. We conclude with some summary comments.
