On Social Security Investment, Or, What About Chile?

With the election over, it’s time to move on to new things, and the folks at the Campaign for America’s Future have asked me to do some writing about Social Security, which sounds like some big fun, so here we are.

We’re going to start with some reasonably simple stuff today, just to get your feet wet; by the time we get a few stories down the road there will be some complicated economic analysis to work through—but let’s begin today by looking a bit south.

Those who support privatizing Social Security in this country often point to Chile as an example we could follow, and that seems like a good place to get the conversation going...so set your personal WayBack Machine to Santiago, May, 1981, and let’s see what we can learn.

“Of what avail are any laws, where money rules
alone,
Where Poverty can never win its cases?
Detractors of the times, who bear the Cynic's scrip,
are known
To often sell the truth, and keep their faces!”

--Ascyltus, from Petronius’ “The Satyricon”

In 1981, Chile adopted a privatized Social Security-like (pension) program that requires most workers to contribute 10% of their income to a private investment account. They may contribute up to 20%. These accounts are maintained by a number of private companies (known as Administradoras de Fondos de Pensiones, or AFPs) that compete for the business by advertising directly to the investing public.

These providers charge commissions and fees for certain services which are paid on top of the contributions.

An additional 3% is collected from most workers for Disability Insurance; 7% more is deducted from wages for health care.

At retirement, the money is either used to purchase an annuity from a private provider to provide a steady source of income or it’s withdrawn at a set rate over time directly from the account.

Those who are self-employed do not have to pay into the system, but they have the option to do so if they’re so inclined.

If you don’t have enough in your private account to purchase an annuity or to withdraw steady amounts over time, but you’ve been contributing for more than 20 years, you will receive a minimum pension from the Chilean Government...but you will also lose any contributions you made to your private account.

AFPs are regulated as to how they may invest; if, through investment losses, they do not have enough money to capitalize the accounts they carry they must provide the money out of their own cash reserves. If they follow the rules, and still lose so many assets they can’t continue to operate, a government bailout is in order.

At the same time, a second “welfare” program (PASIS) was established to create a “safety net” that would provide a benefit of 75% of the poverty level or 25% of your last 10 years’ earnings, whichever is higher.

You can’t collect from both programs, but it is possible to collect from neither. More about that later.

Employers do not contribute to funding the system, however, all employers were forced to give 17% pay raises to their workers to come up with the money for the workers to make their contributions. (Chile was a military dictatorship at the time, making the “forcing” process much easier than it would be in the US today.)

The system is just turning 30 years old, and we’re now seeing the first big wave of workers who are eligible to retire.

So how has all this been working out for Chileans?

The first thing we learn is that the poorest workers probably won’t do well enough to qualify for “top tier” pensions, even though it’s projected that they’ll tend to pay for the benefit over their working lives...which will reduce their income over their working lives. (It’s also projected that workers with higher incomes should do reasonably well.)

Since most workers are poor (Chile has some of the most unequal income distribution on Earth), in the end it’s starting to look like the problems of finding enough money to support the social safety net are actually getting worse, and not better.

Additionally, other problems have come to light:

--You have to find money to “transition” from one system to another, and transition costs have been quite expensive indeed: 6.1% of Gross Domestic Product (GDP; that’s a measure of the total output of an economy) in the 1980s, 4.8% in the 1990s, and 4.3% until 2037. If we were to duplicate the Chilean experience in the US economy, 6% of the 2008 GDP (about $15 trillion) means about $900 billion annually in transition costs for the first ten years, and something north of $600 billion annually for the last 37 years of the exercise.

(Keep in mind that Chile only provides 2/3 of their population with either PASIC or a pension; since we cover a higher number than that in the US, expect those numbers to come in higher than we're guessing here.)

Why are so many not covered? Lots of workers are working outside the “official” economy to avoid making contributions that they won’t get back later (in 1994, it was estimated that only 52% of workers regularly contribute to their accounts); additionally, many women have never participated in the labor force.

--Because the service providers are competing for the business, administrative costs (read: advertising and sales commissions) have been far higher than in the US Social Security system, where administrative costs have been at .07% of distributions, or lower, since 1990. To put this another way, during the 1990s the US Social Security Administration was paying $18.70 per year to administer a claim; at the same time Chile’s various providers were paying an average of $89.10 to do the same thing.

--All that competition, some say, has lead to lots of changing of providers, which tends to make any investment program less efficient over time. (In 1996, half of Chilean workers switched providers; it’s estimated that reduced pension accumulations across the entire system by about 20%.) The Chilean Government made changes in 1997 to try to work through this problem, and they seem to have had some considerable effect.

Evidence suggests most of the switching not related to consolidation in the AFP business is being done by a small percentage of account holders, with some switching as much as eight times in a year; today the average Chilean seems to change AFPs about once every five years. Unemployment also seems to be related to switching; this because the unemployed can establish a new account with a lower set of fees if they move to a new provider.

--Many Chileans, despite living in a system that has, for almost 30 years, required them to manage their own money, actually know very little about that money.

Less than half know that the contribution rate is 10%, only 1/3 know how much (within 20%) is in their accounts, and, according to work done at the University of Chile, “few” actually know what they pay in fees and commissions.

--Those who end up in the welfare program are guaranteed 75% of the poverty level; that suggests that if you’re elderly and on welfare, you’re living in poverty. Because of limited funding, there are qualified elderly poor in Chile who do not receive any benefit.

Today, in the US, about 12% of the elderly live in poverty. Without the current Social Security system in place, it’s estimated that 49.9% of the elderly would have been living in poverty in 2002.

--In Chile, taxes to cover the transition costs tend to rise faster than the “assets under management” for most workers, leaving them less well-off than before—an effect that is most common among the “financially illiterate”...meaning, of course, most Americans. In other words, reform, in Chile, tends to help the wealthiest and best educated at the expense of those who are less of either.

That’s a whole lot of detail, so let’s pull pack and look at the “macro” picture:

Chile has operated a version of a privatized system since 1981, and for the most part the working poor will never see any benefit from the transition. Since Chile doesn’t have much of a middle class, it’s hard to see how the Chilean experience would affect our middle class.

The US Social Security system has reduced the estimated rate of elderly poverty from nearly 50% to roughly 10%; such a reduction in poverty did not occur in Chile with their privatization.

The costs of moving to the same system here, if our experience were the same as Chile’s, would run anywhere from $600-900 billion annually for at least 50 years. Of course, since we provide a Social Security safety net to almost all of our citizens, as opposed to 2/3 of the population, as Chile does, it’s reasonable to assume our costs would be more or less 1/3 higher.

Chile forced its private-sector employers to raise wages to cover the workers’ costs of transition; I’m aware of no proposals that would, or could, impose such a cost on employers in the US.

It appears that Chilean-style privatization encouraged about half the population to engage in “under the table” work, making the funding problem for the system even worse that it would be otherwise.

Frequent switching of account providers is great for the providers, as it creates lots of chances to collect fees for opening and closing accounts and the like—but it’s not so great for the account holders, who are losing up to 20% of their potential earnings more or less because maintaining a sales force and running lots of ads are effective business practices.

It is unknown what happens when a shock like the recent recession hits the system, and we are awaiting research that will help us understand what happens when and if the State is required to refund losses incurred by the AFP if they “follow the rules” but still lose so much money that they lack sufficient capital to operate.

The costs of operating the PASIS program go up even as the cost of operating the retirement accounts are also still high, and the question of whether Chile can continue to expend “safety net” coverage to the 30% of the elderly poor who are not covered remains unknown.

So there you go: there are going to be lots of proposals to privatize Social Security this year, “getting a Chilean” may well be one of the options you hear Conservatives promote—and hopefully by now you have some idea why this doesn’t look like nearly as good an idea as some folks would tell you it is.

Next time, we’ll talk about proposals to invest Social Security money in Treasury debt, and whether such an effort is actually an investment at all.

It’ll be at least medium geeky...and hey, who doesn’t love that?