Becker Friedman Institute
for Research in Economics
The University of Chicago

Research. Insights. Impact. Advancing the Legacy of Chicago Economics.

Can Economic Modeling Take Some of the Guesswork Out of Environmental Policymaking?

In an ongoing series, scholars examine how turbulent policy may move, shake or paralyze the economy

For three days in August 2014, residents of Toledo, Ohio were forced to use bottled water for drinking, brushing teeth, food prep, even bathing children. Cyanobacteria from algal blooms in Lake Erie—blooms fed by phosphorus traced to agriculture run off in the lake’s Western Basin—had poisoned the municipal water supply.

Lake Erie’s problems are dramatic, but they’re hardly unique. Increasingly, communities across the country are finding that their waters are sometimes too slimy, toxic, or smelly to use or enjoy. In 2009, the EPA’s first-ever National Lakes Assessment revealed that more than 20 percent of lakes in the U.S. have high levels of phosphorus or nitrogen. Roughly one out of three U.S. lakes is positive for microcystin, the toxin that closed Toledo’s taps, and in 1 percent of all U.S. lakes, that toxin is present at “levels of concern.”

Worse news for policymakers looking for solutions was the assessment’s report on 35 years of remediation efforts focused on a subgroup of particularly troubled lakes. While 26 percent had improved with intervention, 23 percent had gotten worse. The rest—51 percent—had merely held their own.


Don’t just sit there, do something

It’s fairly easy to defer action on a report. It’s impossible to ignore nearly half a million people without drinking water.

In Ohio, state and local officials hurriedly passed a new law prohibiting area farmers from spreading manure or particulate fertilizers when the ground is frozen or saturated or when heavy rain is forecast. Another law mandated that all the state’s farmers be trained and certified in the proper use of fertilizers by 2017.

The legislative ripples widened when Great Lakes and St. Lawrence governors and premiers met in Quebec in June 2015. Acknowledging that the area’s environmental condition had deteriorated “to the point that it poses a barrier to achieving the economic value and environmental well-being of the entire lake,” they agreed to take necessary actions to reduce the load of “total and dissolved reactive phosphorus” entering the Western Basin by 40 percent by the year 2025.

But exactly what actions are necessary?

Through both direct experimentation and careful modeling, scientists have made real progress in understanding small lake dynamics. Large lake systems, however, present infinitely more unknowns. Their size alone guarantees that they will be subject to a greater range of stressors: nutrient loading, erosion and sedimentation, lake level declines, or invasive species that alter the vertical food web, affecting how nutrients are processed.

The unknowns can make it difficult to extrapolate how known lake processes are likely to play out in a large lake. Will powerful winds and waves that stir up phosphorus-bearing sediments and cause algae to clump have the same effect in a large lake? Will a sudden load of nitrogen feed the production of deadly toxins in the same way? A simple example: In 2015, the second largest algae bloom in recorded history appeared in Lake Erie. By rights, toxin levels should have gone through the roof. Yet concentrations in Toledo remained well below the levels that had kept faucets shut the prior year.

One of the most critical unknowns, for both resource managers and policymakers, concerns whether what is true for small-lake ecosystems even applies to large lakes. Can a large lake, like a small one, reach an “alternative stable state”? That is, can it reach a state of environmental degradation that will be unaffected by incremental interventions like reducing new inputs of phosphorus? Limnologists think so, but they have no way to prove it without risking the loss of an entire large lake ecosystem in the process. If they’re right, then reducing the phosphorus entering Lake Erie’s Western Basin to zero might not be enough to restore the lake’s economic and environmental value.

Meanwhile, decisions and choices must be made. The policymakers who must make them might begin by reviewing an intriguing body of work by University of Wisconsin-Madison economist William Brock, much of it in collaboration with eminent limnologist Stephen Carpenter. Brock’s models evaluate a variety of environmental policy choices through the lens of economic utility, calculating the net present value to society from activities that can pollute our waters (e.g., increased agricultural production or residential and industrial development) and from healthy ecosystems (clean water, fish and game, recreation).

His analyses uncover hidden values, suggest new options for action, and offer more than a few useful guidelines for policymakers threading their way between the rock of scientific uncertainty and the hard place of political exigency.


Mind the lag

Lags are inevitable in policymaking. On the legislative side, coalitions must be formed; constituents need time to adapt; and funding must be identified. On the resource management side, it takes time to choose and implement policies that allow development, agriculture, or aquaculture while preserving lake ecosystems.

Yet Brock and Carpenter’s models show that policy lags themselves have a destabilizing effect on those ecosystems.

Policymakers will do well to realistically evaluate the likely interval between decision and implementation. The research shows that implementing policy gradually—phasing in a planned increase in phosphorus inputs as undeveloped land is rezoned for subdivisions, for example—can compensate for the effect of lags. But if gradual implementation isn’t possible, the bottom line is this: phosphorus input targets should be substantially lower than those derived from traditional deterministic models.


Account for environmental uncertainty

Political uncertainty is not the only condition that should trigger lower phosphorus targets. Whenever phosphorus inputs are unpredictable—heavy rains may or may not wash away nutrients, for example—or decision makers are uncertain about how a lake will respond to altered inputs, then targets must also be reduced.

Analyzing 21 years of detailed data on both phosphorus inputs and the levels of the nutrient observed in Madison’s Lake Mendota, for example, Brock and Carpenter found that the lake’s actual phosphorus levels often exceeded those predicted by a model based on known inputs.One possible explanation is phosphorus recycling: even when inputs are low, wind and wave action stir up phosphorus-laden sediments and keep overall levels of the nutrient high.

Assuming certainty—e.g., that reducing inputs will automatically lower concentrations and improve the health of a lake ecosystem—can lead to policy error. In the case of lakes whose phosphorus concentrations are low at the outset, policymakers who put their faith in this  best-case scenario will set permissible inputs too high. For lakes already at risk, they’ll set those inputs too low, effectively overestimating the economic value of the intervention in the belief that curtailing agriculture or development will restore the lake.

Since uncertainty is unavoidable, the optimal policy maximizes utility over both best- and worse-case scenarios. So if the goal is to maximize economic utility, phosphorus input targets should be lowered. The cost of setting and enforcing those lower targets essentially insures decision makers and their constituents against the risk that the lake will recover slowly or not at all from eutrophication.


Take advantage of early warning signs

Healing a sick lake is expensive. Bringing back a lake that has “flipped”—or undergone an ecological regime shift—is many, many times more expensive, if it’s possible at all.

In 2011, both Nature and Science published the results of studies by Carpenter, Brock, and colleagues that described a set of early warning signals that indicate when a lake ecosystem is approaching the tipping point between “sick” and “possibly terminal.” Principal among those signals is a sudden spike in the variance of measurable phenomena, such as the amount of phosphorus sequestered in the algae that is present in a volume of water.

Given the time it takes to reach consensus on a course of action—and the time it will take for that action to take effect—it just makes sense to keep an eye on the canaries in the coal mine.


Think beyond command and control

Legislation and its accompanying sanctions are not the only tools with which to ensure cooperation. They may not be the most effective, either, particularly in situations of political disequilibrium such as those that characterize environmental issues. A law mandating 50-foot buffer strips between cultivated land and waterways may only harden opposition to government interference; an auction in which farmers can offer that land at a price that compensates them for the loss of its use is more likely to win cooperation.


Economic incentives are anathema to many, particularly with respect to the environment, and most of us would like to think we don’t need financial compensation to do the right thing. But it is human nature to prefer to continue to use and enjoy what we have right now—a green lawn, a strip of income producing land—rather than to give it up for some future and public benefit. Exploring the cost and likely efficacy of alternate economic incentives may help us secure that future benefit before we’ve squeezed out the last penny of profit from the present course of action.


Remember that caution is not the same as confusion—or ignorance

Scientists are professionally cautious. Scientific knowledge is usually derived from models and datasets of multiple events, and so it is expressed as probability, not certainty.

That doesn’t mean we can’t depend on science. Allowing good science to inform public policy—giving weight to credible scientific knowledge and opinion—increases the odds of good decision making in the face of uncertainty.

For more information

For a closer look at the complex interactions of ecosystems and political/social systems

Management of Eutrophication for Lakes Subject to Potentially Irreversible Change, Ecological Applications 9(3), 1999, S. R. Carpenter, D. Ludwig, and W. A. Brock


For a discussion of the signs of tipping points

Early Warnings of Regime Shifts: A Whole-Ecosystem Experiment, Science vol. 332, 2011, S. R. Carpenter, J. J. Cole, M. L. Pace, R. Batt, W. A. Brock, T. Cline, J. Coloso, J. R. Hodgson, J. F. Kitchell, D. A. Seekell, L. Smith, and B. Weidel

Early-warning Signals for Critical Transitions, Nature 461(3), M. Scheffer, J. Bascompte, W. A. Brock, V. Brovkin, S. R. Carpenter, V. Dakos, H. Held, E. H. van Nes, M. Rietkerk, and G. Sugihara.

About this project

This brief and video are produced as part of the Price of Policy Uncertainty, the institute’s research initiative that studies the economic impact of uncertainty.  Produced with the generous support of the MacArthur Foundation, this is the third in a series highlighting the work of leading researchers at the University of Chicago and elsewhere who are exploring key questions in this area. You can see the first video on the challenges of measuring uncertainty here, and the second one on the impact of uncertainty on asset pricing here.