Climate policy as a constrained optimisation problem

Reuters

Constrained optimisation problems are familiar to first-year economics students from the use of indifference curves to solve consumer choice and cost minimisation problems.

But the most commonly used and tractable form of constrained optimisation is linear programming, a mathematical technique for solving logistics problems developed by Leonid Kantorovich for the Soviet Red Army and George Danzig for the US Air Force during World War II to solve military problems.

Kantorovich used it to calculate the optimal distance between supply trucks crossing the frozen Lake Ladoga during the siege of Leningrad in the winter of 1941-42 to ensure they would not sink.

In the post-war world, linear programming has come to be employed in a huge range of industries.

Crude oil refiners use linear programming to determine which crudes to buy and what fuels to produce to maximise their margins. Airlines use it to plan their route networks to maximise revenue and minimise fuel bills. Logistics firms use it to find the best way to schedule parcel deliveries.

Even more complicated optimisations are now possible, including systems that allow preferences and constraints to take non-linear forms. But the basic idea (maximising or minimising an objective function subject to a number of constraints) is the same in each case.

Policy preferences

In the case of energy and climate policy, most voters and policymakers say they want energy to be clean, affordable and reliable/secure. Unfortunately, there tend to be trade-offs among these three goals, and different members of the public express a preference for different combinations.

Environmental campaigners give overwhelming priority to meeting emissions goals. Only after emissions are curbed do they worry about considerations of cost and reliability.

At the other pole, some voters doubt that the climate is changing at all or assign a very low priority to taking steps to deal with it. Voters and politicians in this group assign absolute priority to affordability and reliability.

Most voters and politicians have a mix of preferences somewhere between these two extremes. They want energy to be clean and affordable and reliable. Even so, there is quite a lot of variability in the weights that individual voters and politicians assign to these goals.

Much of the debate about climate and energy policy is framed in terms of the "science," as if there were a scientific basis for determining the right answer or the most efficient solution to the optimisation problem. In fact, most of the differences are not really about the science or evidence about costs but rooted in different preferences about outcomes.

Environmentalists and climate sceptics disagree for much the same reason that conservatives and liberals disagree on questions about income distribution and inequality.

Changing public goals

For the most part, societies deal with this sort of collective choice problem through the political process. Political leaders aggregate the different preferences of individual voters to try to reach some sort of compromise.

The response to climate change is more about political choice than scientific modelling. Models can illustrate the consequences of various different courses of action (and inaction), but they cannot make the choice itself. That decision is inherently political.

But politicians find that the problem with trying to aggregate voters' preferences over affordability, security and sustainability is that the preferences are themselves quite unstable over time.

In the late 1990s and early 2000s, there appeared to be strong support for ambitious policies to curb greenhouse emissions even if the transition to a low-carbon economy proved quite expensive. Recently, voters and policymakers have appeared far more worried about the costs.

Unfortunately, many investments in the energy sector are very capital intensive and have payback periods stretching for years or even decades. Making such long-term commitments in the face of unstable voter and political preferences is inherently difficult.

Policy constraints

The other side of the optimisation problem is the constraints including physical laws, the current state of technology, and a country's natural endowments of resources such as coal, oil, gas, wind, solar and wave energy.

In theory, the constraints should be more stable than voters' preferences. Obviously the laws of physics don't change. But other constraints turn out to be surprisingly flexible and changeable. Technology obviously changes, sometimes quite dramatically. Resource endowments also turn out to be quite variable.

The shale revolution showed that a shift in technology could fundamentally alter the understanding of how much oil and gas might be technically and economically recoverable.

So climate change and energy policy turn out to be a constrained optimisation problem in which both preferences and constraints shift significantly over time.

Moreover, there is some evidence that the preferences and constraints are not independent but influence one another.

When oil and gas resources appeared to be running out, it was much easier to build support for ambitious policies to tackle climate change.

But the shale revolution seems to have changed the balance by holding out the prospect that fossil energy can remain much cheaper.

Finally, both preferences and constraints are surrounded by enormous amounts of uncertainty. Since there is so much uncertainty about future preferences and constraints, the optimum solution is not well specified but embraces a wide range of possible policy choices and may be quite fuzzy.

The constantly shifting nature of both the preferences and the constraints, together with the tremendous amount of uncertainty surrounding both, explains why policymakers find it impossible to sustain a durable consensus around energy and climate policies.

This is an edited extract of an article originally published on Reuters. Reproduced with permission.