Ten days before President Barack Obama took office in 2009, his economic team released an analysis of a proposal to use fiscal policy to stimulate the economy to reduce unemployment. The analysis claimed that the stimulus would create 3.675 million jobs, and that as a result the unemployment rate, which otherwise might rise to nine per cent, would instead remain below eight per cent.
Such calculations appeared to be scientific and precise. However, as it turned out, fiscal stimulus was enacted in 2009, and yet the unemployment rate hit 10 per cent, which was higher than the level that the economists had predicted it would reach without the stimulus. In fact, throughout Obama’s first term, the unemployment rate was higher than the economists projected under the “no‐stimulus” scenario.
The economists were necessarily wrong about one or both of their claims. If they were correct in forecasting that without the stimulus unemployment would top out at nine per cent, then their analysis that the stimulus would generate less unemployment was incorrect, because the opposite happened. Conversely, if the stimulus reduced unemployment below what it otherwise would have been, then it was their baseline forecast of unemployment that was erroneous.
To make judgments about the efficacy of fiscal stimulus in reducing unemployment, we would like to know with reasonable certainty how the 2009 stimulus worked. However, the only scientific way of doing so would be to re‐create economic conditions exactly as they were early in 2009, hold everything else constant, and allow the economy to proceed without enacting any fiscal stimulus. Comparing the results in these otherwise identical economies, with and without fiscal stimulus, would be an application of experimental methods. (Of course, it would be even better if you could perform more than one such experiment. Such replication is another valuable element of the scientific method.)
Evidently, it is impossible to run such controlled experiments in the economy. Instead, what economists do is construct mathematical models and then perform experiments on the models. While such computations can offer apparent numerical precision, there is no evidence that they are reliable.
Ever since World War II, the economics profession has gravitated toward the pseudo‐science of mathematical and statistical modelling. As a result, economists have made policy recommendations in some areas, such as fiscal stimulus, with unwarranted overconfidence. Meanwhile, they have failed to persuade the public of economic insights that are indeed robust.
People who have never taken an economics course, or who have not absorbed the most important lessons from such a course, have economic ideas that are too concrete. They extrapolate from what they experience in their immediate circle, without appreciating the complexity of the economy as a whole.
Academic economists, on the other hand, have drifted into habits of analysis that are too abstract. Their mathematical models also mischaracterize the economy.
In between the too‐concrete economics of the untrained and the too‐abstract economics that predominates in academia, there is what I call Goldilocks economics. Goldilocks economics emphasizes the centrality of the phenomenon of specialization and trade.
In a primitive hunter‐gatherer society, there is very little specialization and trade. Everyone is capable of providing for themselves the food, clothing, and shelter that they need. Knowledge is shared. Food is shared. Each resource must be used no faster than it can be replenished by nature.
A modern economy operates differently. The goods and services that we all consume require millions of different tasks to be performed. If you were on your own, you could provide almost none of them for yourself. By the same token, the work you do probably does not yield anything that you could consume.
Knowledge is highly specialized. Even two people working in adjacent offices within the same organization may not know enough about each others’ tasks to be able to fill in for one another.
Modern societies do not simply consume resources. We also produce tangible resources, including sophisticated tools, durable equipment, and new breeds of plants and animals. Such tangible resources enable us to consume more with a given amount of natural resources. These tangible resources are augmented by intangible resources, including knowledge, innovation, and cultural capital.
If everyone understood the difference between a modern economy and a hunter‐gatherer economy, then perhaps people would be less susceptible to the primitivism of the “sustainability” movement. We certainly want to engage in economic activity in a sustainable way, but this does not mean that we must use only renewable resources.
Modern humans do not merely consume resources. We also produce resources. These include tangible resources, such as highways and tools. They also include intangible resources, such as knowledge and culture.
Because we produce so many tangible and intangible resources, no one natural resource can be tracked as an indicator of sustainability. Instead, the best guess about overall sustainability can be found in market prices. Generally speaking, patterns of specialization and trade that are profitable are sustainable. Patterns of specialization and trade that are not profitable are, according to the market, not sustainable.
Patterns of sustainable specialization and trade (PSST) constantly evolve. Innovations, changes in taste, and events can cause some tasks to become unsustainable and other tasks to become profitable.
Fifty years ago we needed telephone switchboard operators, and today instead we need mobile phone salespersons.
Macroeconomics, which is the branch of economics that purports to connect fiscal stimulus with employment, tries to ignore the evolution of PSST. Interestingly, macroeconomics straddles the too‐concrete thinking of the public and the too‐abstract thinking of the academic elite.
To the general public, Keynesian economics is presented as “spending creates jobs, and jobs creates spending.” From a concrete perspective, this makes intuitive sense. If there were more spending on my product, then my business would hire more workers. If my business hired more workers, then they would spend more.
However, in the economy as a whole, jobs are not created by spending. Jobs are created when entrepreneurs collectively discover new PSST. This takes time, and it takes trial and error. While it would be comforting to believe that government spending works like a gas pedal that can be used to speed up this process, there is no compelling reason to believe that this should be the case.
At the too‐abstract level of economics as encountered in graduate school, nobody teaches that “spending creates jobs and jobs create spending.” However, almost all macroeconomic theory ignores specialization and instead tries to look at the economy as if it were one gigantic GDP factory, producing a single type of output using non‐specialized labour. Academic economists create mathematical models of this GDP factory that purport to explain fluctuations in employment without ever considering the process of creating PSST.
In fact, the economy is not a machine, and it does not have a gas pedal. If we want the rate of job creation to increase, we should look for ways to increase the flexibility of markets and the pace of new business formation. It is by no means obvious that enlarging the role of government will be helpful in that regard.
The scientific pretensions of modern economists, especially macroeconomists, are counterproductive. Economics is a discipline, and I wish that the discipline were better taught and more widely understood. But it is not a science.