If you ask most people why they support raising the minimum wage they’ll says it’s because it helps the poor. But as David Neumark, a scholar at the Federal Reserve Bank of San Francisco notes, numerous studies have shown that there is no statistically significant relationship between raising the minimum wage and reducing poverty.

That finding may appear to be counterintuitive. After all, if poor people have low wages then increasing their wages should help reduce their poverty. To some extent, this is true. However, what is overlookedis that minimum wages target individual workers with low wages, rather than families with low es. The reason that distinction is important is because most workers who earn the minimum wage are in e families.

That es more obvious when you think about position of the American workforce. If you are from a middle-class family, your first job is likely to have paid minimum wage. The same goes for all your friends who are from families higher on the economic ladder. And it’s the same for young workers today. Go down to the mall and you’ll find that the young men and women working in Forever 21 and Abercrombie & Fitch are not from families that are in poverty. Increasing the minimum wage merely ensures that these young people who are (mostly) from wealthier families get a pay raise.

The relationship between being a low-wage worker and being in a e family is fairly weak, as Neumark explains, for three reasons:

First, 57% of poor families with heads of household ages 18–64 have no workers, based on 2014 data from the Current Population Survey (CPS). Second, some workers are poor not because of low wages but because of low hours; for example, CPS data show 46% of poor workers have hourly wages above $10.10, and 36% have hourly wages above $12. And third, many low-wage workers, such as teens, are not in poor families (Lundstrom ing).

Considering these factors, simple calculations suggest that a sizable share of the benefits from raising the minimum wage would not go to poor families. In fact, if wages were simply raised to $10.10 with no changes to the number of jobs or hours, only 18% of the total increase in es would go to poor families, based on 2010–2014 data (Lundstrom ing). The distributional effects look somewhat better at a higher threshold for low e, with 49% of the benefits going to families that have es below twice the poverty line. However, 32% would go to families with es at least three times the poverty line. By this calculation, about a third of the benefits would go to families in the top half of the e distribution.

Moreover, if we consider raising the minimum wage higher, for example to $12, only 15% of the benefits go to poor families, because other higher-wage workers who would benefit are less likely to be poor. Likewise, 35% would go to families with es at least three times the poverty line. With a $15 minimum wage the corresponding figures would be 12% and 38%. This evidence—coupled with the fact that employers who would pay the higher minimum wage are not necessarily those with the highest es, but instead may be owners of small businesses with low profit margins—indicates that minimum wages are a very imprecise way to raise the relative es of the e families.

Some people might claim that even if the minimum wage mostly benefits wealthier families we should still do it since there is some benefit to the poor. But this is another case of failing to consider the difference between what is seen and what is not seen. What is seen is that some poor workers get a pay raise; what is unseen is that many more will simply be unable to find work.

From 2006 to 2012, the average effective minimum wage rose by $1.72 across the United States. Economist Jeffrey Clemens released a paper last month that examines the effect of that wage increase on low-skilled workers:

My baseline estimate is that this period’s minimum wage increases reduced employment among individuals ages 16 to 30 with less than a high school degree by 5.6 percentage points. This amounts to 43 percent of the decline in this group’s employment between 2006 and 2012. Further, it accounts for a 0.49 percentage point decline in the employment to population ratio across all individuals ages 16 to 64.

Not only did the effect of the minimum wage account for nearly half of the unemployment effect on young workers without a high school degree, it lead to nearly a half a percentage point effect on all unemployment.

Keep in mind that this effect was because of a $1.72 increase over a six-year period. Several cities and states plan to raise the minimum wage to $15 an hour—nearly 70 percent more than the current federal minimum wage—over the next two years.

While the change may be a boon to the middle-class kids working at the mall, the effect on low-skilled workers will be devastating. But politicians get elected by doing what is popular (and minimum wage increases are popular) not for doing what is best for the most vulnerable members of our society. Until more people understand how minimum wage laws harm the poor, the kids at the Gap will continue to get more spending money while the poorestworkersstay unemployed.