17 Jul 2009

Using Software to Model Environmental Damage

A guest post by Blake Shurtz*

Economists use stock-and-flow models to forecast the economic impact of environmental changes in the face of uncertainty.

One popular application of the stock-and-flow model is towards climate change. By constructing feedback loops between geophysical and economic parameters we can estimate the optimal discount rate. The discount rate is simply how much we value future damages in the present. The literature abounds with arguments for high and low discount rates between 0.1% and 7.0%. Economists argue for an optimal discount rate on the grounds of efficiency: A rate that is too low would mean overinvestment in emissions reductions while higher return investments in non-climate-related capital are neglected, and vice versa.

I am using the stock-and-flow model to analyze the effects of drought in the Mahbubnagar district of Andhra Pradesh. Given a variety of agricultural parameters (soil condition, rainfall, water availability, slope, elevation, etc.) based on historical data, I am using statistics to predict the availability of water in the future and the economic impacts. It is essentially a risk-assessment: if X water is available, the production losses will cost a farmer $Y. Or to rephrase it: If water costs $X, the farmer can increase their income to $Y by switching to crop Z, which uses less water (holding all other inputs constant).

I am using three pieces of software:
  1. Stata, to predict macroeconomic effects on the secondary and tertiary parts of the economy.

  2. WinEPIC, a non-proprietary, open source, stock-and-flow model available for download form Texas A&M. You can download it here and the Aggies will email you the password.

  3. ArcGIS to perform geostatistical analysis of non-spatial data. ArcGIS lowers transaction costs by bundling the weather and economic information together in a visual framework; see image at right.
Bottom Line: 70% of India’s population practices subsistence agriculture. Agriculture’s share of GDP has been decreasing for years. This is may suggest that development agriculture is becoming more efficient, but there is work to be done. To learn more about stock and flow models, take a class at your university.

* Recent graduate of the Environmental Economics and Policy program at UC Berkeley now doing research at ICRISAT. Read Blake's blog here. ICRISAT is a non-profit and non-political research organization that serves the poorest of the poor in the semi-arid areas of the developing world. We use science as a means to serve the poor, not as an end in itself. This is the human face of science and agricultural research.


  1. WaterSource/WaterBank17 Jul 2009, 13:29:00

    Manair River flows through the study area and joins Godavari River. Mula vagu, Nakka vagu, Pedda vagu etc are the other sub-streams flowing in the study area, which ultimately meets Manair River. All these streams and rivers remained dry due to insufficient rains.

    The average annual rainfall ranges from 900 - 1500 mm.

    Strange that the study used 450 mm.


  2. Blake and Sreedhar23 Jul 2009, 09:49:00


    Although I can not confirm or deny your claim that the average rainfall ranges from 900-1500, I am skeptical for two reasons:

    1. Our data is collected daily, at 3 locations. http://vasatwiki.icrisat.org/index.php/Main_Page
    For the longer time-series data going back 10 years collected by the India Meterological Department, the average rainfall has been far below normal, ranging from 450-500 mm. Where is your data from?

    2. You are making an assertion about the Godvari River Basin while we are focusing on a smaller area within the basin. What area does your data cover?

    I disagree based on my data but I further must ask, based on what you have said here: How is the Godvari River Basin dried up if it was relieving 900-1500 mm of rain annually? If this was true, I would have to change my research topic!


Read this first!

Make sure you copy your comment before submitting because sometimes the system will malfunction and you will lose your comment.

Spam will be deleted.

Comments on older posts must be approved (do not submit twice).

If you're having problems posting, email your comment to me