Heterogeneous construction of scientific knowledge and practice: I. A case of simulating the future of a salt-affected agricultural region

The concept of heterogeneous construction applied to science highlights the ways that scientists mobilize a diversity of resources and, in so doing, engage with a range of social agents.  This idea is illustrated in the next three posts.  In this first post, a situation is described.

The Simulated Future of a Salt-Affected Agricultural Region

In the Kerang region of the southern Australian state of Victoria farmers irrigate some crops as well as pasture, which is grazed by beef or dairy cattle and sheep.  Soil salinization has been a chronic problem; during the middle 1970s, after some very wet years, the problem was acute.  The rise in salinity, following a decline in beef prices, threatened the economic viability of the region.  In late 1977 the Ministry of the state government overseeing water resource issues commissioned the “Institute,” an economic and social research organization in Melbourne, the capital city 240 kilometers south of the Kerang region, to study its economic future.  An agricultural economist from the Ministry and the principal investigator from the Institute formulated a project to evaluate different government policies, such as funding regional drainage systems, reallocating water rights, and raising water charges.  This evaluation would take into account possible changes in the mix of farm enterprises and in farming practices, such as improvements in irrigation layout, drainage, and water management.  The analysis was to be repeated for different macroeconomic scenarios as projected by the Institute’s national forecasting models.

The central part of the project was the construction of what came to be known as the Kerang Farm Model (KFM).  Using a technique called linear programming for each of four composite representative farms, the KFM would determine the mix of farming activities that produced the most income.  Different factors, such as water allocation, could be changed and the effect on the optimal income and mix of activities ascertained.  The division of labor in the project was as follows.  The principal investigator, an econometrician, continued his work on the agricultural component of the Institute’s forecasting model.  The agricultural economist conducted extensive surveys of farm operations for forty farms and acted as liaison with two senior agricultural extension officers in the region who helped to refine the production relationships and parameters used in the KFM.  I was hired for fifteen months as a statistician and modeler to analyze the farm surveys and to construct and operate the KFM.  The Ministry maintained oversight of the project through its agricultural economist and through regular meetings with the project team and an advisory committee.

There were a number of tangible products of the study: the survey and data analysis incorporated in one report to the Ministry; the KFM and economic analysis which made up the second report; a technical monograph documenting the KFM; papers presented at two national conferences of agricultural economists; and a public meeting in the Kerang region to explain the results of the study (Ferguson, Smith, and Taylor 1978, 1979; Taylor 1979).  Even without refinements that were omitted to meet the Ministry’s deadline, detail in the KFM was sufficient to allow evaluation of the required range of factors—the model was complex, yet still manageable to use.

Figure 4.1  A schematic of the Kerang Farm Model (from Taylor 1979).  The arrows correspond to equations that capture the effect of parameters and variables (represented by the boxes) on other variables.

At the public meeting to present the study’s findings some local agricultural extension officers objected to a conclusion that irrigation of pasture gave a better return than irrigation of crops.  This ran contrary to the advice they had been giving to farmers ever since the decline in beef prices.  Reanalysis, incorporating generous increases in crop yields into the KFM’s parameters, was completed rapidly.  This showed that pasture irrigation was a robust result, which could be attributed to the recovery of beef prices by the late 1970s.  The Ministry, meanwhile, focused its attention simply on results that indicated that water charges were not a primary limiting factor on farm enterprises or viability.  Although the Institute had been commissioned to analyze a larger range of options, the other analyses were eclipsed by the conclusion about water charges, which suggests that justifying an increase in water charges had been the Ministry’s primary concern all along.

Readers who have performed contract research on government policy may be all too familiar with the experience of results used for purposes more limited than or in a different spirit than they had hoped.  Cynics would assert that distortion is what one should expect when governments commission research—why should it turn out any other way?  A less cynical, but still fatalist sentiment often expressed by researchers is that the best they can do is to produce results that are as faithful as possible to reality and hope that, eventually, the truth in these analyses will filter through the political process and act as a check on unjustified policy.

An alternative to cynicism and fatalism rests on teasing open the relationship between the scientific research and the circumstances in which it is conducted.  If particular aspects of the relationship can be shown to influence the results in their own way, then one has also shown specific ways the research could have been carried out differently.  Researchers need not wait then for the inevitable distortion of results or for the eventual acceptance of the truth; they can instead attempt to change the particular aspects over which they have most influence.  The possibility of identifying sites where research might be modified motivates the analysis to follow in the next post.

Extracted from Taylor, Peter J. 2005. Unruly Complexity: Ecology, Interpretation, Engagement. U. Chicago Press.


Ferguson, J., A. Smith and P. Taylor (1978). Economic Aspects of the Use of Water Resources in the Kerang Region. Melbourne: Institute of Applied Economic and Social Research, Melbourne, Australia.

Ferguson, J., A. Smith and P. Taylor (1979). Economic Aspects of the Use of Water Resources in the Kerang Region. Melbourne: Institute of Applied Economic and Social Research, Melbourne, Australia.

Taylor, P. J. (1979). The Kerang Farm Model. Melbourne: Institute of Applied Economic and Social Research, Melbourne, Australia.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s