Over the past quarter-century, New York City has experienced two periods of steep economic decline accompanied by fiscal crisis or stress, followed by two extended periods of growth. In both economic crisis periods (1970-77 and 1989-93) the city's fiscal problems were compounded by rising debt burdens which forced the city to set aside larger shares of shrinking or stagnant budgets for debt service payments. During the two economic expansion periods, however, the city has taken different paths in terms of debt management. Over the 1978-88 recovery and growth period, New York City sharply reduced the mountain of debt it had inherited from the fiscal crisis. In the current recovery and growth period (dating from 1994), the city's debt burden has become heavier relative to ability-to-pay.

• The city reduced total debt outstanding from 19.3 percent of fiscal year personal income and 196.4 percent of city tax revenues in 1978 to 11.9 percent of personal income and 134.9 percent of tax revenues in 1988. Over this period, debt service payments also fell from 30.9 percent of tax revenues to 12.2 percent.
• As the city began to emerge from its last recession in 1993, total debt outstanding was back up to 13.5 percent of personal income and 151.2 percent of tax revenues, while debt service payments were 13.9 percent of tax revenues. As of 1997, four years into the recovery, total debt outstanding was still 12.9 percent of personal income and had risen to 164.9 percent of tax revenues. Netting out the large pre-payment of 1998 debt service, debt service absorbed 15.8 percent of total tax revenues. By 2002, total GO, MAC, and TFA debt is forecast to decline slightly further to 11.7 percent of personal income but rise further to 172.4 percent of total tax revenues; debt service payments will climb to 19.2 percent of tax revenues.

Several factors underlie these different trends. During the 1980s expansion, average annual debt growth (3.7 percent) was held below inflation (6.4 percent), while average tax revenue growth (7.6 percent) was above inflation but under average personal income growth (8.8 percent). (Thus in real terms, average personal income growth was 2.2 percent, average tax revenue growth was 1.1 percent, and average debt growth was -2.6 percent.)

Over the current and forecast expansion period, in contrast, average debt growth (3.9 percent) exceeds inflation (2.7 percent) while average tax revenue growth (2.4 percent) is below inflation; both lag average personal income growth (5.6 percent) by a wide margin. (In real terms, average debt growth is 1.2 percent, average tax revenue growth is -0.3 percent, and average personal income growth is 2.8 percent.)

All this clearly reflects the current administration's emphasis on tax reduction, as well as some continued underlying weakness in baseline property and overall tax revenue growth. The point that needs to be made here is that such high debt burdens so deep into an economic expansion means that the city's capital program is especially vulnerable to recession. If the music stops on Wall Street, city tax collections could drop precipitously, and New York City would almost immediately confront debt service ratios well over 20 percent of total tax revenues-or be forced to make very deep cuts in its capital program to hold down debt service costs.

In the past few years, New York City has used parts of the large budget surpluses generated by the Wall Street boom to prepay debt service. But such large prepayments cannot permanently cushion us against rising debt pressures. A more lasting response-one that would reduce our exposure to the risk of capital retrenchments when the rainy day finally comes-would be to use parts of the current surpluses to actually reduce debt, either by retiring existing debt or substituting pay-as-you go financing for some future debt.

The Transitional Finance Authority was created to allow New York City to continue funding a capital program that was suddenly being constricted by the city's constitutional general debt limits. This is a temporary solution. In devising a permanent one, three issues must be addressed:

• On what (if any) basis should the city's debt-incurring power be limited?
• How large should the city's debt-incurring capacity be?
• How should city debt be treated in budgetary and accounting terms?

Currently, the city's general debt-incurring power is limited to ten percent of a moving five-year average of full valuations of taxable real estate. Lagged changes in the full valuation average have created huge "whiplash" effects in the city's debt limit. The GO debt limitation rose from 108 percent of tax revenues (9.8 percent of personal income) in 1981 to 208 percent of tax revenues (17.9 percent of personal income) in 1990 to 306 percent of tax revenues (27.2 percent of personal income) in 1994-and then fell back to 165 percent of tax revenues (13.0 percent of personal income) in 1997, and will drop even further to 140-145 percent of tax revenues (about 10 percent of personal income) over the 1999-2002 plan period.

TFA's establishment flowed, in part, from the observation that the recent sharp drop in the ten percent limitation created an illusion of capital financing improvidence where in fact there was none. Actual and forecast growth in GO debt averaging only 2.9 percent over the 1994-2002 period hardly seems profligate-until it is matched up with negative 4.4 percent growth in the general debt limitation over these same years.

But the deception works the other way as well. Between 1988 and 1994, GO indebtedness more than doubled (from $10.1 billion to $23.1 billion), growing at an average rate of 14.8 percent-but the city's debt-incurring power almost tripled (from $19.9 billion to $55.4 billion), averaging 18.6 percent growth. The precipitous growth in the city's debt-incurring margin over these years-it went from $3.2 billion in 1988 to $19.7 billion in 1994-did not evoke an outcry for constitutional reform. But it was just as much a symptom of the problem as was the subsequent rapid decline in the margin.

As could be seen from comparing the general debt limitation to total tax revenues, the underlying problem here is that the year-to-year changes in the property tax base do not reflect the city's true overall capacity to support debt. In part this is due to the way property is assessed, but the larger problem is that the property tax makes up a shrinking share of the city's total tax base. Forty years ago, property taxes made up 67 percent of total city taxes; today, property taxes account for 36 percent. It seems clear that the debt limitation should be calculated from a much broader tax base.

The city's debt-incurring power limit can be reformed either by adding to the existing base or by replacing some of the existing base. The city has proposed legislation to do the former. This would base the city's debt limit on "ten per centum of the average full valuation of taxable real estate plus ten per centum of the average aggregate personal income of residents."

IBO estimates that this would nearly double the city's currently forecast debt limit over the plan period, for example raising the 2002 limit from almost $32 billion (9.9 percent of personal income, 146 percent of total tax revenues) to about $62 billion (19.4 percent of personal income, 286 percent of total tax revenues). In real terms, this would bring the city's general debt-incurring power back up close to its 1993 and 1994 peaks.

But it should be noted that even after the sharp drop in the New York City's debt limit since 1994, the city still actually has, in both real and relative terms, a higher GO debt-incurring power today than it did at any time before 1989 (at least going back to before the fiscal crisis). This is not to deny that a somewhat enlarged debt limitation could be prudent and appropriate. But it is difficult to make the case that doubling the current limitation would be. Even if "only" two-thirds of the proposed taxable real estate-plus-aggregate personal income limitation was used, this would itself send debt service costs spiraling well above 20 percent of tax revenues.

Over the past two decades, through all its ups and downs, the city's general debt limitation has averaged about 175 percent of total tax revenues (14.6 percent of personal income). This might be taken as a rough indicator of where a long-run sustainable debt limitation, based on the city's whole tax base, could be set. It would enlarge the city's current general debt-incurring power by about a fourth.

Accounting for debt

The city's proposed legislation also includes provisions granting the city very broad powers to set up finance trusts-authorities-that can issue bonds secured by "irrevocable" rights to city taxes and revenues (and by finance trust property). These authorities would be controlled the city, but their debt would not be considered city debt.

Two observations are in order here. First, the proposed legislation contemplates establishing the power to create authorities in addition to vastly expanding the city's general debt-incurring power. Taken together, these two provisions seem to open the door for virtually unlimited expansion of debt supported by city taxpayers. Things would be different if, for example, the taxes and revenues used to secure authority debt were subtracted from the GO debt limitation base. But the legislation does not include such a restriction.

A government financial statement should enable creditors, legislative and oversight officials, and the general public to accurately assess the full scope and costs of government activities. The 1999 executive budget removes the debt service of the Transitional Finance Authority (TFA) and the personal income tax revenues dedicated to paying that debt service from the primary city budget. Consequently, the city's debt burden is made to look smaller than it really is, and the ratio of city debt to revenues seems to level off over the next four years. In reality, the portion of city tax dollars being absorbed by debt will continue to climb.

Removing TFA provides a skewed and incomplete picture of the allocation of resources by city government. The executive budget shows larger shares of total city spending going to areas such as health and public safety, and smaller shares going to debt service, than would actually be the case if the budget were adopted as proposed.

If the city's ability to create off-budget authorities is expanded as the legislation proposes, the ability of the public to monitor city indebtedness would become increasingly difficult.