Yesterday on his blog the President’s Budget Director, Peter Orszag, asks himself and then answers the question, “How much does the federal government owe?”
This sounds like a technical question of concern only to “those of us wearing the green eyeshades,” but the Director’s suggested answer has dangerous ramifications, and could mislead, or at least confuse taxpayers and financial market participants.
The Director’s answer makes the federal debt appear $1.4 trillion smaller than the way it is traditionally measured. He argues that we should, in effect, ignore 1.4 million million dollars borrowed by the federal government. That is breathtaking.
Let’s look at the Director’s argument and why I think it’s dangerous.
Most budget experts focus on debt held by the public, which Director Orszag accurately describes as “the amount that the federal government owes to others.” I will expand on that a bit with some concrete numbers:
- Take the total amount the Federal government will spend this year. Specifically, we’re looking at cash”paid” by the U.S. government to someone outside the government in 2009. A budget wonk would call these outlays. I’ll use the nonpartisan Congressional Budget Office’s numbers for current law, so I get $3.85 trillion of outlays for 2009. That is way (way) above historic norms, in part due to the financial stabilization efforts, and in part due to the new “stimulus” law.
- Now take the amount the Federal government will collect in revenues this year. This is cash coming into the U.S. government from someone outside it. Almost all of this is taxes. CBO says this is $2.186 trillion of revenues for 2009.
- If the U.S. government is paying out $3.85 trillion in cash (outlays) this year, but collecting “only” <sigh> $2.186 trillion in cash, then we need to come up with the difference somewhere. That difference is $1.667 trillion for 2009. This is what CBO says is the federal budget deficit for 2009.
- The U.S. government gets this cash by issuing IOUs to people outside the government, aka Treasury bonds. The government gets cash from anyone who buys Treasury bonds – individuals, firms, and foreign governments.
- The debt held by the public is simply the accumulation of these IOUs. It is the sum of money owed by the U.S. government to others. (Update: See the caveat at the bottom.)
Nothing I have said so far is the slightest bit controversial, but this is where Director Orszag and I part ways. Tuesday he wrote:
As I said at the beginning of this post, I think the most meaningful measure of federal debt is debt held by the public net of financial assets. If I take a $100 loan from my bank and stick that amount into my bank account without spending any of it, my family and I aren’t poorer, because even as I owe $100 to my bank, my bank owes $100 to me. On net, and as long as the new asset is equal in value to the new liability, there’s no change in my overall financial state. There’s a similar effect when the federal government borrows money in order to invest in financial assets.
Suppose I tweak the Director’s metaphor to make it better fit the current situation and illustrate my point. If he takes a $100 loan from his bank and invests it in the business of his deadbeat neighbor Alan I. Gorp, he still owes the bank $100. The bank cannot loan that $100 to anyone else. His (the government’s) borrowing has “crowded out” borrowing by someone else. And who knows how much his $100 investment will be worth next month? We should care not just about his net position, but also about his total liabilities, and especially about how much he (the government) is borrowing from the bank (private sector).
In normal times this would not be a big difference, because the U.S. government in large part stays away from owning financial assets. Now, however, the federal government is buying equity stakes in banks and other large financial firms, and issuing loans to financial and non-financial firms. Director Orszag’s numbers show that the U.S. government owned $506 billion of financial assets last year, and will buy another $915 billion this year. (I’m subtracting “Debt net of financial assets” from “Debt held by the public” on Table S-1 of the President’s budget.) Those are huge numbers, and have a huge effect on what figure you cite for the federal debt.
If you look at the traditional measure of debt held by the public, which you’ll remember is the sum of all IOUs (Treasury bonds) issued by the Federal government, then under the President’s budget and using OMB numbers, that’s equal to $8.36 trillion. Compared to one year of our entire national output (GDP), that’s almost 59% of GDP.
If, however, you net out OMB’s estimate of the value of the financial assets, then the debt held by the public net of financial assets, is “only” $6.94 trillion, equivalent to almost 49% of GDP. That’s still a big bad number, but it’s $1.4 trillion and 10% of GDP less bad than the debt held by the public numbers. That’s a convenient way to make the problem look much smaller. Director Orszag argues that it is also the “most meaningful measure of current federal debt.”
Here is his key paragraph:
As the federal government has acted to stabilize the financial sector amidst the worst financial crisis since the Great Depression, the federal government has purchased significant financial assets … such as preferred equity stakes in Fannie Mae and Freddie Mac. The federal government will likely take a loss on these purchases, but the assets have value. And just as what my bank owes me should be netted against what I owe the bank in determining the health of my personal finances, the value of these assets should be netted against publicly held debt in determining the health of the government’s finances. … Debt held by the public net of financial assets is the most meaningful measure of current federal debt …” (emphasis added)
I disagree with this last statement, but I think I understand why he says it. From his perspective of the federal budget, he’s netting out some of his liabilities with a somewhat liquid asset that he now holds and hopes someday to sell. He concedes the point, however, that he is including some assets and liabilities with his new measure, but excluding others. This makes his new metric suspect.
From the perspective of the U.S. economy, the “netting” comes from different places. The U.S. Treasury has to issue $905 billion of Treasury bonds this year to raise the cash to buy those financial assets. This makes it harder for private firms and individuals to borrow, because they are competing with the government for cash, so they have to pay a higher interest rate. Those funds are then invested in other parts of the economy.
Another way to see why this is a poor metric is to imagine that the U.S. government were to borrow another trillion dollars by issuing even more Treasuries, and then immediately buy one trillion dollars of credit default swaps with the cash raised. According to Director Orszag’s preferred measure, nothing would have changed, because the two transactions would net out. But clearly we would have just had a major impact on the U.S. (and global) financial economies. U.S. government borrowing in these enormous amounts hurts financial markets, no matter what is done with the funds raised.
Director Orszag touches on another problem with his new metric when he writes “The federal government will likely take a loss on these purchases, but the assets have value.” He’s right, but the value of the particular assets being purchased by the government is highly uncertain. How much is he counting as the value of the $19.4 B loaned (so far) to General Motors? I sure hope he is not counting it at face value. What about the $70 B “invested” in AIG, or the $5.5 B in Chrysler? Any private firm valuing these assets would say their values need to be discounted.
The values of these financial assets are highly uncertain and depend heavily on what assumptions OMB uses about the likelihood of them being repaid. For people to trust this metric, they need to understand how it is calculated, which means that OMB should divulge the discounts they are applying to their financial assets. I will guess that he does not want to divulge those assumptions. I wouldn’t if I had his job.
I think the most meaningful measure of current federal debt is still debt held by the public. I think the public policy debate can be further informed by also disclosing the estimated value of the financial assets held by the U.S. government. But policymakers should not net out the two and use that measure instead of the one that most directly measures how much the U.S. government is borrowing from the private sector. This is particularly true when that new measure hides $1.4 trillion of debt borrowed by the U.S. government from the private sector.
Director Orszag, and those measuring his performance, should continue to use debt held by the public as the most meaningful measure of current federal debt. Budget projections will account for that measure to come down over time as the financial assets are sold and funds recouped.
Net measures can hide meaningful information. This is a theme I will return to often. Any time someone in economic policy gives you a net figure, see if you can learn something more by asking about the components that make up the net calculation.
The President’s Budget is titled “A New Era of Responsibility.” In his February 24th Address to the Congress, the President said,
The only way this century will be another American century is if we confront … the mountain of debt they stand to inherit. That is our responsibility.
A new era of responsibility does not begin with hiding $1.4 trillion of that mountain of debt. These IOU’s will not go away just because we ignore them.
Update (12:20 PM Wed): A friend corrects my statement that the debt is simply the accumulation of past deficits. It’s not. The Credit Reform Act measures credit subsidies (like for federal loan or loan guarantee programs) differently than it measures cash flows, and the deficit does not capture “means of financing and cash management, like when Treasury borrows funds and deposits the cash at the Fed.” I stand corrected on these points. I don’t think this changes my logic above about whether to net out the purchase or sale of financial assets.