There are two types of ideas that dominate current public discourse—the contrarian and the counterintuitive. A contrarian idea is one that, whether correct or incorrect, opposes or rejects popular opinion or goes against current practice. A counterintuitive idea is one that is contrary to intuition or mon-sense expectation but is nevertheless correct. Getting the two mixed up can have a detrimental effect on society.

Take, for example, the increasingly popular contrarian-posing-as-counterintuitive idea that locking up more criminal offenders isn’t making people any safer. As the Washington Post‘s Emily Badger writes,

As economists would put it, there are diminishing returns to incarceration. Lock up one criminal in town, and crime will decline. Lock away two, and it will probably decline further. But each criminal in prison yields a smaller and smaller impact outside of it — until finally, there’s no new impact at all. Now we’re effectively imprisoning more and more people with no benefit to public safety.

The first four sentences are perfectly reasonable, but the last sentence draws the wrong conclusion. Let’s create a simple model to show why that reasoning is flawed.

Imagine a remote island—Theft posed of approximately 1,000 men. Last year on the island there were exactly 100 thefts. Because of a peculiar genetic anomaly in the region, individual criminals in Gangland are only able mit one theft per year. No one is able to leave e to the island, there are no women (hence no new islanders), and no one has died or will die in the next five years.

Based on these facts, we can know that since there were 100 thefts last year there were 100 thieves. The local police surmise that for every thief they lock up, the theft rate will go down by one. But something strange happened. In year one they locked up 30 criminals and the theft rate dropped to 70. In year two they locked up 20 criminals and the rate dropped to 50. In year three, though, they lock up 15 additional criminals. The police assumed the theft rate would continue to drop linearly, but instead of falling to 35 the theft rate was 40.

Now imagine a social scientist on Theft Island claims to know what happened. “We’re effectively imprisoning more and more people with no benefit,” he says. “Locking up additional criminals will have no effect on the theft rate so we should reduce the number of people who are incarcerated or release some of the thieves now in prison.”

How would you respond to that social scientist? Let’s ask the 19th century French economic journalist Frédéric Bastiat. In the opening to his famed essay, “What Is Seen and What Is Not Seen,” he says:

In the department of economy, an act, a habit, an institution, a law, gives birth not only to an effect, but to a series of effects. Of these effects, the first only is immediate; it manifests itself simultaneously with its cause – it is seen. The others unfold in succession – they are not seen: it is well for us, if they are foreseen. Between a good and a bad economist this constitutes the whole difference – the one takes account of the visible effect; the other takes account both of the effects which are seen, and also of those which it is necessary to foresee.

What is seen (at least initially) on Theft Island is the effect of locking up criminals: crime goes down. But what is not seen is what the effect would be of not locking up criminals or releasing those who were currently incarcerated.

One of the factors that remains unseen is the reason the thefts stopped declining in a linear fashion. Perhaps some former non-criminals became thieves. But what should be forseen is what would happen if you stopped locking up thieves or let them out of prison: they would have mitted another theft in the future, thus causing the crime rate to rise even higher.

Now let’s apply this to the real world. If you incarcerate a criminal, what happens? Well, for starters, they are unable mit a crime against the public. A thief in jail isn’t able to steal your car stereo. If you put a criminal in jail then, as Bastiat might say, the number of crimes they would mitted drops to zero. That is what can be foreseen. As Inimai M. Chettiar notes, “Criminologists call this the ‘incapacitation’ effect: Removing someone from society prevents them mitting crimes.”

Yet Chettiar isn’t convinced this effect is sufficience:

What do the numbers say? Did this explosion in incarceration cause the crime decline?

It turns out that increased incarceration had a much more limited effect on crime than popularly thought. We find that this growth in incarceration was responsible for approximately 5 percent of the drop in crime in the 1990s. (This could vary from 0 to 10 percent.) Since then, however, increases in incarceration have had essentially zero effect on crime. The positive returns are gone. That means the colossal number of Americans cycling in and out of prisons and jails over the last 13 years was not responsible for any meaningful fraction of the drop in crime.

Let’s assume these statistics are accurate. What could be the reason that increased incarceration is not having as large an increase on the crime rate? Here are three possibilities:

1. The One-Time Effect — Criminals who got caught and put in jail were not going mit any more crimes in the future. Locking them up has no effect on future crime rates since they wouldn’t mitted new crimes.

2. The Substitution Effect — The number of active criminals at any period in time is relatively stable. So if you lock one up, another person—who was previously not inclined to criminality—will take their place. Locking them up has no effect on future crime rates since they are simply replaced by new criminals.

Neither of those two seems all that plausible, so let’s look at a third possibility:

3. The Extra-Criminal Effect — A subset of mit a disproportionate number of crimes. Locking them up reduces the crime rate significantly for as long as they are incarcerated.

The extra-criminal effect can explain why the “positive returns are gone.” If you lock up a regular criminal, you are only reducing the future crime rate by a small amount. But if you lock up an extra-criminal (or someone who would have e a extra-criminal) you affect the crime rate substantially.

Let’s also assume that it’s no easier to catch extra-criminals than regular criminals and that they are randomly dispersed throughout the criminal population. What happens if there is an increase in incarcerations? There is an increased likelihood of catching the extra-criminals, and thus a greater likelihood of drastically lowering future crime rates.

The “incarceration” effect has the most robust impact on crime rates when something is done (e.g., increasing the number of incarcerations) to remove those mit a disproportionate number of crimes. This is so obvious that you almost have to be a social scientist to miss the point. Yet many criminologists go even further and claim that if we stopped putting people in jail that it would not lead to an increase in the number of mitted. In essence, they deny what they already know—that putting people in jail prevents them mitting crimes—in order to advocate for their preferred political agenda.

If we are putting people in jail that have mitted any crimes, then that is a grave injustice and should be immediately rectified. But if we are incarcerating people that mitted crimes and would continue to do so until they got caught, then it would be insane to empty the jails because the “incarceration” effect is no longer having an exponential effect.

Segregating criminals from the public isn’t the only solution to reducing crime. But let’s not go full-contrarian and believe that increased incarceration of the guilty has had no effect on public safety. mon-sense solutions really do make the most sense.