In his speech presenting his economic agenda Texas Gov. Rick Perry included this among his ideas on Social Security: “[R]eturn to pre-1983 law and allow state and local governments to newly opt out of Social Security and instead allow their employees to pay solely into state or locally run retirement programs. This has been done around the country, with better results. We ought to allow it again.”

He’s not the only one to toy with the idea of devolving Social Security to the states. Newt Gingrich added his voice to those cheering for defederalization of Social Security at the debate on Saturday night: “Here in Texas, there is the Galveston system, where they discovered that if you put in about half as much money in the private sector you would get twice as much as you would giving it to government.”

But does this make sense? There are many reasons to think not, which is why none of the bipartisan commissions has given it serious consideration. In fact, they’ve suggested the opposite, as the New York Times reported: “The proposal runs counter to the recommendations made recently by a couple of high-profile, bipartisan debt reduction commissions — which called for shoring up Social Security, in part, by enrolling the nearly seven million state and local government workers who are not in the program.”

There are three essential problems with taking employees out of federal Social Security. It’s not necessarily good for employees. It doesn’t address the central problem we face. And it would be huge burden on state and local governments.

To begin with, from the employee point of view, putting retirement funds in the hands of states is a risky proposition:

Some city workers in Central Falls, R.I., were never enrolled in Social Security, which saved both them and the city from the payroll tax that funds the program. The city filed for bankruptcy this summer and reduced their pensions by as much as half, to as little as $10,000 a year. Now they have no Social Security benefits to fall back on.

Public schoolteachers in Illinois do not participate in Social Security, and their pension fund is in precarious shape. Last year the fund reported having just 48 cents for every dollar of benefits it had promised, largely because the state had failed to pay the required amount of money into it for many years.

Lawmakers in Maine recently considered a proposal to enroll more of its government workers in Social Security, but discovered that the transition costs would be crushing.

There are also concerns about fairness and logistics here. The right to pull out of the Social Security system belongs to the employer. If you want to stay within the Social Security but your employer does not, too bad. What about employees who go from a non-Social Security employer to one that participates in the program? As for a better return for employees, it’s not at all clear that Gingrich is right. The Times reports: “Studies of the Galveston plan have determined that it provides a better deal for some high-income workers, but a worse one for low- and middle-income workers. And its benefits are not protected from inflation, as Social Security’s are.”

Looking at this idea from the perspective of the long-term solvency of Social Security, the Perry-Gingrich idea moves us in the opposite direction. We should be adding bodies:

Enrolling those workers in Social Security would have another benefit, the panel found: the payroll taxes that would be collected would close 8 percent of Social Security’s projected 75-year shortfall. Social Security’s last crisis was resolved in part by bringing the federal work force into the program in 1984.

Adding better-paid state and local workers like teachers and police officers, who make up most nonparticipants today, would give the federal system’s solvency a bigger boost than might be expected. This is because Social Security is set up to shift money toward lower-income people. The better-paid newcomers to the program would tend to add more taxes during their working years, but would not draw out a commensurate amount when they retired.

Then there are the states to consider. It turns out Galveston isn’t such a great example. The Times explains that “the Galveston program is actually more expensive than Social Security. Social Security’s contribution rate is 12.4 percent of payroll, with 6.2 percent coming from each worker and 6.2 percent coming from the employer. The Galveston plan’s total contribution rate is 13.9 percent of the payroll, with 6.1 percent coming from each worker, and 7.8 percent from the county — or, more specifically, its taxpayers.”

In his critique of the idea of devolving Social Security to the states, Mitt Romney argued that it would place a crushing financial burden on the states. Looking at four states, a campaign white paper argued:

• Florida could face a $16 billion shortfall . . .

• Iowa could face a $1.5 billion shortfall — more than 25% of its general fund.

• Arizona could face a nearly $3 billion shortfall — more than 33% of its general fund.

• Louisiana could face a $1 billion shortfall — more than 10% of its general fund.

If this is accurate, few states would want to take on the burden.

So what really is the point of sending Social Security back to the states and localities? There isn’t much of one. We’d been further weakening the federal system and sending workers into an array of plans that would lack the financial guarantee of the federal government. We’d create a maze for employees moving from one job to another or one state to another.

This, it seems, is the sort of “big idea” that complies with a stringent view of the 10th Amendment ( Perry at times has argued that Social Security is unconstitutional) but entirely fails in the real world.

Bold can be positive when it comes to reform. But in this case Gingrich and Perry have a lot of explaining to do if they want their idea to be treated as a realistic proposal.