Saturday, April 15, 2017

CBO on government debt and its budget impact


CBO reported that the federal budget deficit rose $63 billion in the first half of fiscal 2017 (OctoberMarch) to $522 billion from a year earlier. Those aren’t estimates; they are actual numbers. But, they were used to introduce CBO estimates of deficits over the next 10 years. Since the forecast has implications, it's worth looking at how much value to place on CBO estimates. Only then does it make sense to look at the substance of their forecasts.

You gotta love the CBO.

While Obama was doubling the national debt over eight years, the CBO wasn't just low-profile on the issue; in many cases, they actually obfuscated it. For example, they grossly underestimated the deficit impact of the stimulus plan. If you think it was a partisan ploy, you might note that this was done by overestimating its stimulative impact. As I argued in a bunch of postings in September 2010, every neo-Keynesian was overestimating the stimulative impact because of flaws in their models. CBO was just particularly bad.

If you need another example of CBO's wacko analysis, the Affordable Care Act (ACA) was estimated to reduce federal deficits by $124 billion by 2019 (March 2010 estimate by the CBO before the law was enacted). They not only got the magnitude wrong; they may have gotten the direction of the impact wrong. Their estimates were so indefensible that CBO has stopped estimating the effect on the deficit from the ACA.

It isn't just economic impacts CBO screws up. In February 2013, CBO predicted that ACA enrollment in the individual market would be 13 million in 2015, 24 million in 2016 and 26 million in 2017. The actual enrollments were 11 million, 12 million and about 10 million. As recently as March 2016, CBO was projecting an enrollment of 15 million for this year. They were equally flaky on their forecasts for Medicaid.

Similarly, CBO was also badly wrong about the 2003 Medicare prescription drug benefit. The drug benefit cost about 40% less over its first decade than CBO projected. You may recall that this was the unfunded entitlement that was going to destroy the world.

The less you think it is a partisan issue, Ryan Lizza in “The Mandate Memo: How Obama Changed His Mind,” (, March 26, 2012) reports that Obama became so frustrated with the CBO that at one point during the healthcare debate he banned aides from using the term “CBO” in his presence. Instead, he called the CBO “banana.”

Don't misunderstand my comment about nonpartisan. I'm not naïve enough to believe that public employees don't respond to the conflict of interest of being a public employee when evaluating the impact of public sector activities. I'm just pointing out that CBO isn't a serious analytical organization. One doesn't need partisan politics to be incompetent. CBO could be incompetent and partisan, not just incompetent, but that's a different issue.

The debt is real and the problem is real

There is one aspect of CBO’s analysis of the deficit outlook that is free of the conflict of interest noted above. To some extent, their projection of the deficit is based upon their assessment of another government organization’s impact, rather than the impact on the public sector of the private sector. Part of their projection for growth in the deficit is an assessment of how much the Federal Reserve (FRB) is going to contribute to the deficit.

The CBO deficit forecasts are predicated on their own economic growth forecasts as well as stability in policies. Policies will change, and the CBO is no better at economic forecasting than anyone else. So, the forecasts themselves are of less value than their analysis of what's driving the debt and deficit. In some instances, the impact of CBO forecasts and assumptions are easily isolated. The impact of the debt on future deficits is one such variable. The impact of the debt is a product of the size of the debt and the interest-rate the government pays.

A logical question is how sensitive are the forecasts to the CBO's assumptions regarding interest rates. CBO estimates that if interest rates are one percentage point higher than in its current projections, the result will be $160 billion additional spending in each year over the next decade. Given that level of sensitivity and the uncertainties associated with any interest-rate forecast, current data on the debt and deficit are important to understanding the forecasts.

The Federal Government’s net interest payments increased $7 billion in March from a year earlier. To put that in perspective, the president’s 2018 preliminary budget proposes cuts of $2.7 billion in discretionary spending. So, the interest costs increase in one month was almost 3 times what the president was able to propose cutting in his first year of operations. Further, it's worth noting that the increase is about 30% for the month. Payments increased by $28 billion for the six months of fiscal 2017 to $152 billion. That is about a 22% increase and it is among the biggest single spending items in the CBO estimates.

The increases reflect the larger debt.  As mentioned, the debt has doubled over the last eight years, but the more ominous fact is that the growth in the debt accelerated in the last year: Federal spending exceeded revenue by $176.2 billion last month. The budget gap was about $68.2 billion higher than a year ago. For the first six months of the fiscal year the deficit was about 15% higher than the same period a year earlier. Over the past 12 months, the deficit stood at $651.5 billion, compared to $460.6 billion a year ago. Those are Treasury Department figures not CBO estimates.

The size and accelerating growth of the debt obviously has implications, but those implications depend upon one's forecast of interest rates that will be applied to that debt.  For the last eight years the FRB’s near-zero policy kept Federal government interest rates at historic lows. That reduced net interest payments even as the overall debt increased. However, at the same time, the federal fiscal outlook deteriorated due to ever-higher future interest obligations. In the CBO forecast the Federal Reserve’s decision to raise interest rates after years of near-zero rates compounds the impact of the debt.

Raising interest rates is not the only way the FRB is influencing the deficit forecast. The FRB’s bond-buying programs earned money that the FRB turned over to Treasury each year, reducing the size of the federal budget deficit by tens of billions of dollars. FRB officials are indicating that this year they may stop buying new bonds as those on its balance sheet come due. That will reduce the number of bonds they have earning interest, and, as a consequence, there will be smaller FRB contributions to the federal budget. Also, as interest rates rise, the FRB will have to pay more on bank reserves deposited at the Central Bank. The FRB pays banks 1% on reserve balances. As the amount the FRB has to pay banks increases, it will reduce the amount they can turn over to the government. The amounts aren’t negligible; they pay about $20 billion a year on reserves.

Thus, in addition to the higher interest rates that the federal government will have to pay, there will be a reduction in the more than $90 billion the FRB has turned over to Treasury in recent years. According to CBO, all of this is set to explode on President Trump’s watch. One might ask why CBO ignored this time bomb for so long.  After all this is not the first time they've done 10 year forecasts, and even they aren’t naïve enough to have assumed that interest rates would never rise. But the more important point is that the chickens have come home to roost, here-and-now, according to CBO.

The FRB is not the only instance where the government pays interest which is then recovers. According to the Flow of Funds data from the Federal Reserve, about 28% of the federal debt is held on other federal accounts. That includes about 16% of the total federal debt held by the Social Security Trust Fund.

Consequently, between the FRB which holds 13 to 14% of the debt and the other agencies that hold about 28%, a significant portion of the federal debt, about 42%, is held in accounts were some or all of the interest is recovered by the government.  Even that 42% estimate understates the total public-sector ownership of federal government debt. Another 3 – 4% is held by state and local governments and about 1% is held by the pensions of state and local governments.

The fact that close to half the federal government debt is held by the public sector determines the implication of the rise in the deficit and debt. Basically, the higher interest payments remove control of how taxes are spent from the current elected federal officials. Some of it is shifted to the state and local governments that hold the debt. But, the main effect is that it shifts those revenues to mandated expenditures resulting from the actions of previous elected officials.

The interest on debt run up by previous administrations and previous commitments to programs that where expenditures are determined by a formula, independent of revenue, consume a larger portion of tax revenue. Social Security and Medicare are the largest such programs, but now they must compete with Affordable Care Act mandated expenditures. The fact that the interest cost of the debt rose by three times as much in one month as the administration was able to cut in its plan for a year of operations illustrates the loss of control implied. Basically, elections become irrelevant as current elected officials have less and less control over government priorities. Increasingly, the only option open to an elected official who wants to control government priorities is to ignore the deficit, run up the debt, and let the next administration worry about it. That is a terrible set of incentives.

No comments:

Post a Comment