For many of us, the Great Recession is not just a distant memory. Economic crises have lingering effects on our bank accounts and our peace of mind. One concern we often have when the economy tanks is that crime will also increase. Yet strangely, in 2008 and beyond, crime actually decreased – quite significantly – in the United States, leaving the nation’s crime experts mystified.

The idea that a weak economy or a rise in unemployment will result in an increase in crime is based on logic and early criminological ideas, suggesting that crime happens when we are denied what we want or need. Interestingly, the decrease in crime when the economy suffered a collapse in 2008 is fairly well representative of the empirical research on the matter.

Research on the economy and crime spans decades and has looked at the relationship from numerous angles. Does crime increase when the economy is weak? Some research finds that it does; other studies show that the direction of this relationship is the opposite. For example, when the economy was booming in the 1960s, crime – serious, violent crime – began to steadily increase. How can we make sense of such a finding?

Despite decades of research trying to reconcile these inconsistent findings, criminologists and other specialists are still puzzled. However, one important part of the economy has been overlooked. Economists call this the “shadow economy.”

The shadow economy is the surprisingly large segment of the economy that is underground. It is the otherwise legal work that goes unreported to the government and is, consequently, not factored into gross domestic product or employment figures. Common examples of under-the-table work, where it is easy to conceal income from the government, include baby-sitting, housecleaning, web design, landscaping and repair work.

Economists have investigated the shadow economy, exploring its size and trends both in the United States and abroad. Estimates of the size of the shadow economy illustrate how important it is; some research has found that in the U.S. alone, the shadow economy is roughly $2 trillion in size, more than 10 percent of U.S. GDP. Clearly, this sector is a non-trivial factor that research should consider when examining the effect of economic swings.

So, what if all of the research exploring how the economy affects criminal behavior is incomplete, because it considers only a partial measure of the economy? It stands to reason that unemployment would not have as large of an impact on crime if there is an alternative way to have needs met.

In a recent study, we examined this question. We considered the possibility that the link between unemployment and crime would be less strong in places and times when there is a robust shadow economy. In other words, do unemployed individuals choose to enter the shadow economy to supplement their income rather than resort to criminal activity? That is precisely what we found. In statistical terms, the size of the shadow economy dampened the effect of unemployment on crime. In short, as unemployment increased, criminal activity (particularly property crimes such as burglary and theft) increased, but that relationship was attenuated in the presence of a strong shadow economy.

This finding has important implications for how we think about the economy’s effect on human behavior. For decades, criminologists have wondered if the logic under which some have operated – that threatening withdrawal of resources will lead to humans seeking alternative outlets for those resources – holds. Our results suggest that it may. That is, humans react to a reduction of resources in predictable ways. But policies such as those popular in the 1960s that seek to reduce crime by improving the legitimate economy are not likely to have the desired effect if the shadow economy is ignored.

The economy is, unsurprisingly, more complicated than criminologists have recognized for years. The shadow economy is an important factor that neither theorists nor policymakers should overlook.