Human beings dislike losses. In fact, they dislike losses a lot more than they like equivalent gains.

This simple point helps to explain what kinds of economic incentives are most likely to have an impact. It also places a bright spotlight on an overlooked obstacle to fiscal reform.

Let’s start with some evidence. Many people have been interested in giving teachers an economic incentive to teach better, by telling them that if their students improve their test scores, they will get some extra money.

Unfortunately, the record is pretty mixed; if teachers are promised bonuses, students don’t seem to do a lot better.

But consider an ingenious study by Harvard economist Roland Fryer and his colleagues, tweaking the usual way of creating incentives. Instead of promising teachers a bonus, researchers gave teachers the money in advance, and told them if their students didn’t improve, they would have to give it back. The result? Student math scores shot up.

Why does a threatened loss of an advance payment have such a big effect, when teachers aren’t much influenced by the promise of a bonus? The answer lies in one of the central findings of behavioral economics, which goes by the unlovely name of “loss aversion.” In short, the prospect of a loss focuses the human mind. Even if people don’t care much about gains, they will work hard, and possibly fight, to avoid comparable losses.

Research shows that even professional golfers display loss aversion. They do significantly better when putting to save par than when putting to make a birdie. Indeed, neuroscientists have found that loss aversion is wired into the human brain.

An appreciation of loss aversion offers some important lessons for public policy. Consider recent efforts to decrease people’s use of grocery bags. In the District of Columbia, one approach was to offer a 5 cent bonus to customers who brought reusable bags. It had essentially no effect.

Starting in 2010, the district tried another approach, which is to impose a 5 cent tax on those who ask for a grocery bag. Five cents isn’t a lot of money, but many people don’t want to pay it. The new approach, enlisting loss aversion, has had a major effect in reducing the use of grocery bags.

The point is that if a company or a government wants to discourage behavior, a fee is likely to have a much bigger effect than an equivalent reward. Even nominal fees can have a big impact. It follows that a subsidy, including a tax credit, is likely to be far less effective than a tax – and that even a pretty large economic subsidy may have a weaker impact than a smaller tax.

If companies or governments want to change human behavior with a new surcharge, they will get a lot more opposition than if they enlist bonuses or discounts instead. The less effective approach is also the more popular one.

We should now be able to see the role of loss aversion in the current U.S. fiscal debates. Any effort to raise revenue, or to cut programs, will impose losses. Consider, for example, tax deductions for mortgage interest, charitable contributions, state taxes and retirement savings. Imagine that these deductions didn’t now exist, and the question was solemnly asked: In the current economic environment, should we create them in their current form?

I am not denying that the right answer might be yes, but the whole discussion would be different from what it is today, when many people are resisting losses from the status quo. Loss aversion helps to entrench existing programs, making them seem like entitlements, simply because they establish the reference point against which losses and gains are measured.

Can anything be done to overcome loss aversion? There are three possibilities.

First, we might try to convince people that however painful, some losses are justified.

Second, we might make loss aversion less relevant by bundling two or more reforms, ensuring that those who lose in one area gain in another.

Third, we might focus people’s attention on the question, not whether we should eliminate or scale back a program, but whether we would now create it in the first instance.

That question has the important advantage of ensuring that if we keep it, we do so because it is a good idea – and not just because of the intense desires of those who resist losses from the status quo.