Welfare programs pay more than minimum wage in 35 states.

That's according to a new study released this week by the Cato Institute, a Washington-based libertarian think tank. It's an update from its 1995 study that examined the same issues.

Its conclusion this time around, accounting for the changes in the government's 126 separate programs for low-income people, is that government aid can be more than the earnings from a regular, entry-level job. And the pay gap has increased in recent years, the study concludes.

Here are some of its numbers:

Not only do government-assistance programs for the unemployed pay more than minimum wage in 35 states, but they also pay more than a $15-an-hour job, according to the report. Hawaii has the "most generous benefit package," following by the District of Columbia and Massachusetts.

In 11 states, these programs pay more annually than the average teacher after his or her first year on the job. In 39 states, it pays more than a starting salary of a secretary. And the comparisons continue.

In total, the federal government spends $668.2 billion on these programs annually, while states give out another $284 billion, the report finds.

Cato's conclusion? Well, the study tries to prove what the institute and other conservatives and libertarians have argued for years:

If Congress and state legislatures are serious about reducing welfare dependence and rewarding work, they should consider strengthening welfare work requirements, removing exemptions, and narrowing the definition of work.

By making it harder to qualify for these programs and adding more eligibility requirements from the updated 1996 Temporary Assistance for Needy Families law, states can help bridge this gap, the study says.

And raising the minimum wage, as President Obama has suggested, is a nonstarter, according to the institute, which argues it raises unemployment for the lowest-skilled workers.

In the U.S., more than 100 million people get some sort of welfare assistance from the federal government, according to a 2012 report from the conservative magazine The Weekly Standard. That number does not include those who only receive Social Security or Medicare.

Cutting off benefits could have a deep impact on those families, many of which are minority or immigrant households. Welfare benefits are also capped after a certain amount of time, which obviously doesn't go for minimum wage.

Food stamps, housing, medical, and other government-assistance programs are often discussed by these groups and have been the target of budget cuts from congressional Republicans. In the House-passed farm bill last month, food stamps were left completely out in order to help its passage. The Democratic-controlled Senate is not likely to pass that bill.

UPDATE

Sharon Parrott, a vice president for Center of Budget and Policy Priorities, a Washington-based think tank, responded to the Cato report, saying the results were misleading. She tells National Journal:

They got the comparison between working and not-working really skewed. The first thing that they do is they assume that people who aren't working have ready access to a large set of benefits that virtually no one receives all of. So, that really exaggerates the benefit package of most families with kids whose parents are out of work. The second thing they do is they ignore the kinds of help that is provided and is available for working families, including families who were not working and receiving welfare and transition to work. And they assume that as soon as they transition to a low-wage job, all of those kinds of supports are immediately terminated, which is also false. So when you exaggerate the benefits available to people who don't work and you dramatically understate the benefits that are available to people who work, it's not surprising that the comparison, or the break-even point, is dramatically skewed in their report. 

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.