Better economic maps – Econlib

• • Our new model includes several innovative features: for example, instead of using a representative household, it features a demographically accurate synthetic population with millions of households (matched by age, education, race and consumption habits). Instead of using a representative company, we model the behavior of tens of thousands of the largest companies, in one-on-one correspondence with real companies….
    • • J. Doyne Boer, Sense of Chaos: a better economy for a better world P. 258

MMainstream economics is willing to build a model of an economy in terms of a ‘representative individual’. One hypothetical consumer represents each household. One hypothetical company represents each company. Many different types of workers are lumped together as ‘labor’. Many different types of machines and other productivity-enhancing factors (such as company reputation or process knowledge) are lumped together as ‘capital’. I have long questioned this way of doing economics, which I call the ‘GDP factory’ method of analysis.

J. Doyne Farmer and his relatively small group of like-minded researchers have advocated for and implemented a different approach for decades. Drawing from the field of ecology, they aim to build models that include agents that employ different strategies within the overall system.

The representative-individual approach involves carefully choosing a set of assumptions in the economist’s mind about human behavior, representing them as equations, and solving the equations for a single equilibrium. It predates the age of computers.

The farmer’s approach, illustrated in his new book Making sense of chaos, requires a very different modeling strategy, called ‘agent-based modeling’. It begins with observations about how different individuals choose strategies to make a living, consume, and invest. The goal is to see how these strategies interact over time. This requires computer simulation.

Take the stock market for example. The ‘representative individual’ approach assumes a single investor with complete information and a single strategy for maximizing return relative to risk. The farmer’s approach instead starts by looking at the types of strategies that different investors actually use. Some focus on the fundamentals. Others try to spot trends. Everyone has different information and uses different heuristics.

The representative-individual models of the stock market tend to have dynamic properties that are uninteresting and unrealistic. They predict minimal market movements, much less trading than we observe, and nothing like the pattern of run-ups and crashes that seems to characterize existing markets. The models with heterogeneous investors are able to replicate the patterns we actually observe in the stock market.

One of the most interesting findings from representative agent models is that as the influence of players using one strategy increases, the dynamics of the financial market change. Strategies that dampen volatility for a while can suddenly cause instability.

For example, Farmer points out that major investment banks adopted value at risk (VaR) as a strategy to manage market exposure in the late 1990s. VaR measures the loss resulting from, for example, an unfavorable price movement of two standard deviations. Using such a metric, a risk manager would say that you can increase risk exposure as market volatility decreases, and that you should decrease it as volatility increases. In good times you get a self-reinforcing feedback loop that pushes asset prices higher as banks expand their portfolios. But then a small setback causes everyone using VaR to try to sell at once, causing a very serious, self-reinforcing loop on the other side. Farmer says this describes what happened in the financial markets before and during the 2008 financial crisis.

Farmers and colleagues have also used computer simulations of heterogeneous agent strategies to analyze the energy market, with a particular emphasis on trying to assess the feasibility of an ‘energy transition’ to prevent climate change. According to their analysis, the main cost of the shift to renewable energy sources is upgrading the electricity grid. But in fact, energy production will be cheaper, so a faster energy transition is generally positive for the economy.

• For example, in 2050, our estimated global annual electricity grid expenditure for the Rapid Transition is approximately $670 billion per year, compared to $530 billion per year for the No Transition. However, the projected total system costs in 2050 are approximately $5.9 trillion for the Rapid Transition and $6.3 trillion per year for the No Transition. While the additional $140 billion in network costs may seem expensive, it is significantly less than the savings that come from cheaper energy. P.253

The mainstream approach to doing economic theory will always have the advantage of being easy to communicate and replicate. When someone shows the results of a mainstream model, you can solve the equations yourself and get an idea of ​​what drives the results.

For empirical work, replication is not that reliable. Farmer reports that when he worked at a company interested in exploiting the inefficiencies of the stock market, his team looked at published articles about market anomalies.

• For about half of the papers, we could not reproduce the results even when we tested the postulated efficiency deviation with the same data. P.146

Simulations are more opaque to those of us who are not part of the team that built the model. We cannot reproduce the results for ourselves. If economists want to adopt agent-based models, they will need to develop ways to articulate, explain, and justify the choices they make in constructing the models.

I think of economic models as maps. If an old-fashioned triptych map said I had to take the George Washington Bridge to get from where I live to Boston, I’d be stuck with that. Using the map on my smartphone, I can consider alternatives and even make adjustments in real time based on traffic conditions.

Enormous amounts of data are becoming available to economists. Computing power has increased by orders of magnitude. This presumably causes the trends to favor agent-based models over the representative-individual standard.

But as maps for policymakers, agent-based models are still far from reliable. I would be careful not to assume that they make centralized decision-making a good way to keep an economy functioning. You should not bet the farm on Farmer.