Advertisement

Analysis: U.S. budget approaching the wall

By MARTIN HUTCHINSON, UPI Business and Economics Editor

WASHINGTON, March 28 (UPI) -- As a former investment banker, I never bought former Treasury Secretary Robert Rubin's self-seeking theory that budget deficits hit interest rates. In a capital market the size of the United States, both a substantial budget deficit and the needs of private industry can easily be financed, without noticeable adverse effect -- after all, retail mortgage originations in 2001 totaled $2.03 trillion, up from $1.02 trillion in 2000. Thus a $300 billion budget deficit is a relatively small item in a debt market that through securitization and derivatives is now wholly integrated.

Nevertheless, as budget deficits surge and keep surging, there comes a point at which they start to matter.

Advertisement

That point is probably well above $500 billion. A $500 billion budget deficit would represent a substantial chunk of the debt market, and certainly strains would be seen if the current $500 billion trade deficit had diminished, thus lessening the surge of foreign capital into the United States. However, such strains would be caused largely by the United States' wholly inadequate savings rate, which prevents the richest country in the world from financing its own economic needs. If interest rates rose, and financing became difficult to come by, the rising budget deficit would not as such be the reason for it, and such policies as a dividend tax cut, which would presumably increase savings rates, would be helpful in solving the problem even if they did marginally increase the deficit.

Advertisement

The relationship between the size of the federal deficit and the "crowding out" effects in the capital markets is not linear. Instead, it is an example of a typical relationship seen in financial markets, in which there is almost no effect over a considerable range, then slowly an effect begins to appear, before it explodes hyperbolically into a "wall" at which in this case the budget deficit becomes unfinanceable. You can see examples all over the place of countries hitting the wall; Argentina in 2001 is the most recent, but Russia in 1998 and Mexico in 1994 are both well-known recent examples.

Close to the wall, the cost of financing escalates rapidly, with quite small movements in the deficit or debt ratio producing large changes in the cost of finance. Generally, at this point, austerity programs are ineffective, because the cost of finance is rising so rapidly that debt costs overwhelm budget savings.

The position of the "wall" depends crucially on the level of a country's debt outstanding before it gets into difficulty. Here the United States is lucky in that its debt to gross domestic product ratio is currently quite low at 60 percent (including the nominal 26 percent held by the Social Security trust fund.) However, it must be borne in mind that from 2018, the surplus accruing in the Social Security trust fund becomes a deficit, and that a huge proportion of the liabilities in the housing market (those of Fannie Mae and Freddie Mac) have a pseudo-guarantee of the U.S. government. Hence, with the federal deficit already running at a fairly elevated level and GDP showing no great signs of growth, the U.S. debt/GDP ratio will inevitably increase sharply over the next few years.

Advertisement

As I remarked above, with a relatively low debt/GDP ratio, it is likely that a federal deficit of 5 percent of GDP ($523 billion, based on 2002's GDP) would be readily financeable, without a significant increase in interest rates -- a number of countries, notably France and Germany seem likely to run deficits of around this size in 2003, and Japan's fiscal deficit has run at 6 percent to 8 percent of GDP for a number of years, without any apparent adverse effect on interest rates.

Conversely, a federal budget deficit of 10 percent of GDP ($1.046 trillion) would be very difficult to finance on a cost-effective basis. Such a deficit has only occurred in the middle of major wars (for example, in 1943-45), when savings can be commandeered for the war effort. Thus in Britain, for example, financing difficulties occurred and interest rates rose sharply when net public sector borrowing approached 7 percent to 8 percent of GDP, in 1980-81 and again in 1992-93.

Hence, if the U.S. federal deficit approaches $1 trillion, trouble is in store, and interest rate rises and "crowding out" can be expected to be severe.

State budgets complicate matters somewhat. According to the National Association of State Budget Officers, state budget shortfalls are expected to reach $29 billion in the year to June 2003 and $82 billion in the year to June 2004. In most states, however, there is a requirement for the state budget to be balanced; while this requirement is honored more in the breach than in the observance, it would appear that a practical maximum exists on the total of state budget deficits, probably of around 2 percent of GDP, or $208 billion. Hence the wall at which the federal budget deficit becomes unfinanceable is probably not at 10 percent of GDP, but at around 8 percent of GDP, or $837 billion. Adverse interest rate rise and "crowding out" effects would become visible at a level lower than this, but probably not all that much lower, maybe at around 6 percent of GDP ($628 billion.)

Advertisement

When I began writing about the federal budget in the summer of 2001, it was in substantial surplus, and I was far indeed from the media consensus in predicting a deficit of $200 billion in the year to September 2002. In the event, given the modest economic recovery in the first half of 2002, I was a little high; the deficit ended at $158 billion -- or about $168 billion if you adjust for timing differences in October 2001.

This year, the deficit in the five months to February 2003 was $194 million compared to $68 billion ($78 billion on an adjusted basis) last year. The Office of Management and Budget is now projecting a deficit of $304 billion for the full year to September 2003, plus $75 billion which will result from the Iraq war. My own prediction last July (excluding any Iraq war) was $350 billion. Given that the first five months of fiscal 2003 were $116 billion (adjusted) worse than the comparable period of fiscal 2002, I think the OMB prediction is still a little low; annualizing the deterioration from 2002 to 2003 gives a federal budget deficit of $437 billion plus the cost of the Iraq war. In practice, I think the budget deficit in the year to September 2003 will stay below $500 billion, even including the cost of Iraq, but not by much.

Advertisement

So what of the following year? Well, it depends on what the economy does. Since I believe that the economy will be in or close to recession for the whole of the period to September 2004 (it may dip and then bounce, or hover boringly close to the flatline, but the net effect is the same) then the deterioration in the fiscal balance in 2003-2004 can be expected to be close to the fiscal deterioration in 2001-2002 ($286 billion) and that projected above for 2002-2003 ($278 billion or a little less, excluding Iraq.) Of course, it is to be hoped that the Iraq war will have ended by September 2003, and that the budget costs of Iraqi reconstruction will be moderate.

But in that case, on, I repeat, my guess as to the U.S. economy's behavior not that of the U.S. government, the federal deficit in the year to September 2004 would be at least $250 billion above the non-Iraq deficit to September 2003, or of the order of $687 billion.

In other words, by September 2004, even though U.S. public debt will still be quite low, the federal budget deficit will have topped 6 percent of GDP, and be getting fairly close to the invisible "wall." At that point, and not before, the gloomy Rubin predictions of increased interest rates and "crowding out" of private sector borrowers may begin to come true.

Advertisement

Even if Rubin may eventually be correct on the prognostication, however, he and his Democratic ex-colleagues are far from reality on the diagnosis. Federal spending growth, which had averaged just over 3 percent per annum under both Democratic and Republican presidents and Congresses from 1991-1999, stepped up significantly to 5.1 percent in the year to September 2000, 4.1 percent in the year to September 2001, and then leaped to 7.9 percent in the year to September 2002. Since the recession did not under any estimate begin until September 2000, the rise in public spending preceded it.

As I have documented earlier, its principal cause appears to have been the replacement of Newt Gingrich by Dennis Hastert as speaker of the House of Representatives. Under Gingrich, final federal budget spending each year was around $30 billion below the president's initial budget proposal($34 billion in the year to September 1998 and $31 billion in the year to September 1999, passed into law in October 1998, for example). Under Hastert, final spending was about $25 billion above the inital proposal in each year. Of course, as spending increases get built into the following year's baseline budget, it is very easy in the latter case for spending to begin rising much faster than GDP, and much faster than revenues even without a tax cut. Since 2001, the combination of weak fiscal control in the House of Representatives, a mild recession and a national security crisis has caused federal outlays to exceed the president's budget by much greater amounts -- $49 billion in the year to September 2002 and probably close to $100 billion including Iraq costs in the year to September 2003.

Advertisement

Since federal receipts in 2002 were the same percentage of GDP as in the recession year of 1991, the problem is not here. Federal spending cannot continue to grow faster than GDP if fiscal health is to be maintained, yet on present trends, as outlined above, serious bond market pressures may start to appear sometime in 2004.

There is another problem. Economic growth in OECD countries correlates fairly well both with the level of government spending and with its rate of change -- about 50 percent of variations in economic growth between countries and between periods can be explained by this means. Contrary to popular superstition, therefore, a "Keynesian" boost in government spending, such as occurred with the 2002 "stimulus package" is likely in all but the shortest run to depress rather than increase growth in the economy as a whole -- thus, of course making the budget deficit worse. Just as "dynamic scoring" needs to be adopted for tax cuts, to account for their growth-inducing effects, so it needs to be adopted for spending increases, to account for their growth-suppressing effects.

If the federal deficit in 2004 is over $600 billion, and interest rates start to spiral, then even though Robert Rubin may have been proved correct -- finally -- on the adverse effect of budget deficits, I wouldn't bring him back to run the U.S. economy because of the extra federal spending he would favor. Instead, I'd rather bring back Newt Gingrich. Save $30 billion from the president's budget each year, building that saving into the next year's baseline, and pretty soon, the problem will have been reduced to manageable size.

Advertisement

Latest Headlines

Advertisement

Trending Stories

Advertisement

Follow Us

Advertisement