Thursday, May 14, 2009

Robert Reich: I was a trustee for Social Security and Medicare

Robert Reich blogs:

What are we to make of yesterday's report from the trustees of the Social Security and Medicare trust funds that Social Security will run out of assets in 2037, four years sooner than previously forecast, and Medicare’s hospital fund will be exhausted by 2017, two years earlier than predicted a year ago?

Reports of these two funds' demise are not new. Fifteen years ago, when I was a trustee of the Social Security and the Medicare trust funds (which meant, essentially, that I and a few others met periodically with the official actuary of the funds, received his report, asked a few questions, and signed some papers) both funds were supposedly in trouble. But as I learned, the timing and magnitude of the trouble depended a great deal on what assumptions the actuary used in his models. As I recall, he then assumed that the economy would grow by about 2.6 percent a year over the next seventy-five years. But go back into American history all the way to the Civil War -- including the Great Depression and the severe depressions of the late 19th century -- and the economy's average annual growth is closer to 3 percent. Use a 3 percent assumption and Social Security is flush for the next seventy-five years.
The problem with Reich's analysis is two-fold. The first problem is that for the economy as a whole, the real social security problem occurs in 2016 (or sooner) and not 2037. Thus, even if we assume 3% growth rather than 2.6% , this is too short a time horizon for the power of compounding to kick in, versus a time horizon out to the year 2037 or beyond.

Why is 2016 critical?

As I have pointed out before, 2016 is when Social Security cash flow reverses. From inflow the fund begins to show a net outflow. The fund will not be out of assets (though they are dubious Treasury security assets), but instead of being a net buyer of Treasury securities, it will become a net seller of Treasury securities. The social security fund currently buys approximately 25% of all Treasury securities issued. That is going to completely stop by 2016.

That is the big problem for the government. Social security is going to be buying fewer and fewer Treasury securities in the years ahead, and then completely stops in 2016, and becomes a net liquidator. This is going to put huge pressure on rates, coupled with climbing inflation and the fact that China is slowing its accumulation of Treasury securities. Owning bonds is not the place to be.

1 comment:

  1. Part of the reason why it seems that Treasuries held by the Social Security Trust Fund have value is that normally when Treasuries are held they have value to whoever holds them even if the Treasury has spent the proceeds from issuing them. However, when the Treasury holds Treasuries it is holding its own debt. In order to spend the value of the Treasuries to pay Social Security recipients, it must sell its own debt which is borrowing. Because of this, the only way in which it could have avoided needing to borrow in order to spend the value of the Treasuries in the Trust Fund would be to have not already spent the proceeds from the Treasuries.

    If the Trust Fund were to withdraw Treasuries from the account, it would have to sell them for cash to give to Social Security recipients. When it is the Fed that sells Treasuries, it is not borrowing. Instead it is sterilization since the Treasury does not get additional funds to spend. However if the Treasury were to sell previously issued Treasuries, it would be the same as if they were issuing new Treasuries since they would be removing the ability to spend from the rest of the economy to transfer to Social Security recipients. This is the same as what would need to be done if there were no savings from previous years in the Trust Fund. It is different than if the Treasury issued the Treasuries and they were then bought and resold later since in such a case the reselling would be offset by the buying. If the Treasuries in the Trust Fund were resold, it would be using one Treasury to borrow twice since they would never have been bought.

    In 2018 when the Social Security Trust Fund will need to borrow it will be in the same situation as if someone were spending more in a year than he was earning that year and needed to borrow because of this. However, he would only need to borrow if he didn’t have savings from previous years, which indicates that the Trust Fund really doesn’t have savings from previous years. The only way in which it can be said that the Trust Fund has savings from previous years is in the same way that someone could say that he had a savings account if it was a savings account where if he wanted to withdraw $100.00 the bank would say that he had to pay them $100.00. Without any savings, he would have to borrow the $100.00 to give to the bank to get his $100.00. He certainly wouldn’t be able to use such savings account as collateral for a loan!

    If government borrowing is increased to replace the increased amount of government borrowing that goes to Social Security, it will harm the economy by increasing interest rates. To the extent that the additional borrowing causes the dollar to fall, interest rates will rise even more. In addition, as the employed percent of the population becomes smaller, GDP will decline in proportion to spending. This will cause the dollar to fall even more. As inflation increases because of these factors, Social Security obligations will increase to the extent that they are indexed to inflation.

    The lack of funding of Social Security will become more apparent in 2011 when baby boomers start to retire. Since the government has depended on having more paid into the Trust Fund each year than is paid out and is using the net inflow of funds for financing spending, even though there will still be a surplus until 2018 the government will have to increase borrowing before then to maintain the same level of spending.