Interest Rates

and Government Debt

Some linkages and consequences
Vito Tanzi and Mark S. Lutz

I, .1 a much-cited address in August 1984 at This article surveys various issues associ- investment is productive, the debt would pay
the 40th Congress of the International ated with the growth of public debt among for itself by raising the level of economic ac-
Institute of Public Finance, Jacques de the Organization for Economic Co-operation tivity and, therefore, the tax base.
Larosiere, then Managing Director of the IMF, and Development (OECD) countries. It high- If the increase in public spending is not
expressed his concern about the "explosion of lights the links between increases in the ratio temporary and is financed by debt, the share
public debt" that was affecting developing of government debt to GDP and increases in of public debt in national income may grow.
and developed countries alike. He outlined real interest rates, between increases in inter- As a consequence, the cost of servicing the
various consequences of excessive indebted- est payments and reductions in capital spend- debt will also grow, especially if interest rates
ness and called on governments to pursue ing, and between increases in interest pay- are high. Thus, ironically, the debt that may
structural reforms in their public finances to ments and increases in taxation. have been used to keep tax rates down may
contain such an "explosion" and put their fis- eventually force these rates up, since the
cal houses in order. Reasons for concern country will in time need to generate a "pri-
Given the high and growing public debts in Debt financing substitutes for taxation. It mary surplus" (the difference between rev-
most industrial countries in recent years, the thus allows governments to maintain or in- enues and non-interest expenditures) to ser-
renewal of such concerns is not unjustified. crease, at least temporarily, public spending vice the debt. In cases where public spending
There has been much controversy, in particu- without the need to legislate tax increases. In is difficult to cut, that primary surplus must
lar, as to whether an enlarged public debt other words, it has the political advantage of be generated mostly through higher taxes.
could raise interest rates. While earlier stud- generating an immediate benefit (the public Over the 1975-87 period, for example, in spite
ies have examined the effects of fiscal deficits expenditure) without an immediate cost (the of the increase in the share of public debt
and debt on real interest rates for individual raising of tax rates). Given that governments into GDP, the share of total taxes in GDP
countries, there is a growing consensus that, are likely to have short horizons and, thus, to grew substantially in most OECD countries,
in view of the increasing integration of finan- discount future costs at high rates—espe- largely to finance higher government interest
cial markets, which allows fiscal deficits of cially when there is a high probability that payments.
one country to be financed by savings of these costs will be faced by another set of pol- The growth of the ratio of debt to GDP is
other countries, a better understanding might icymakers, or even by another party—the influenced by the rate of growth of the econ-
be obtained by taking a global, rather than a temptation to finance spending through debt omy, the interest rate on the debt, and the size
national, perspective. Thus, each country's is naturally strong. of the primary surplus. Thus, a high growth
deficit has its impact on interest rates in the If the additional spending is temporary, an rate and a low rate of interest will be major
world capital market. argument can be made that debt financing contributors to restraining the growth of pub-
will help smooth the required changes in tax lic debt in GDP. The current fiscal stance, as
rates over time. Since sudden and temporary assessed through the primary surplus or
changes in tax rates generate distortions and, deficit, also plays an important role. In fact,
thus, welfare losses, with adverse effects on since countries generally cannot, over the
growth, the use of debt finance will increase longer-term horizon, significantly influence
For a detailed version, see "Interest Rates and the efficiency of the economy. This would be the rates of interest and the growth of the
Government Debt: Are the Linkages Global Rather the case especially when the increase in economy, the policy variable available to
than National?" by Vito Tanzi and Mark Lutz, IMF spending is caused by wars, depressions, or them is the primary surplus. If non-interest
Working Paper (WP/91/6), available from the large public investments concentrated in a spending cannot be reduced, then the tax rate
authors. relatively short time. In the latter case, if the becomes the basic policy instrument.

30 Finance & Development / December 1991

©International Monetary Fund. Not for Redistribution

 In the absence of comparable real effective a specific country will not have any effect on
(i.e., inflation and growth-adjusted) interest Chart 1 the real interest rate of that country. Rather,
rates on the debt, comparisons of debt ratios General government each country's financial policy, apart from its
over time or across countries are not particu- gross dept, 1970—87 impact on the world interest rate, influences
larly meaningful. In recent years, because of (Percentage of gross domestic product) the difference between its interest rate and
large capital movements, interest rates have the average international interest rate.
tended to converge across countries. This im- How can we capture the global link be-
plies that global factors, as distinguished tween world real interest rates and levels of
from purely domestic factors, have played a global government debt? The global real in-
growing role in influencing domestic interest terest rate can be modeled as being deter-
rates. However, even if real world interest mined by the process equilibrating saving,
rates tend to be influenced by the size of the both public and private, and investment. The
global debt, and if countries with higher than demand for capital goods is seen to be
average debt tend to pay some differential strongly procyclical, rising rapidly during pe-
over the international interest rates, then riods of economic expansion and falling dur-
cross-sectional comparisons still need to be ing recessions, and, therefore, may be cap-
qualified. tured by a proxy variable reflecting the
business cycle. Saving decisions are made by
Public debt in OECD countries both the public and private sector. Therefore,
The ratio of general government gross debt measures of government deficits and private
to GDP has been rising in most OECD coun- saving levels, both expressed as ratios to out-
tries, reaching very high levels by 1987 in put, were included in the model. The model
many of them. Between 1975 and 1987, the also tested the ratio of government debt to
ratio rose from 61.1 percent to 132.4 percent GDP, as this may have a direct influence on
in Belgium; from 11.9 percent to 57.2 percent real interest rates, and other variables captur-
in Denmark; from 64.3 percent to 128.7 per- ing government consumption levels and mon-
cent in Ireland; from 60.4 percent to 92.7 per- etary policy.
cent in Italy; and from 22.4 percent to 76 per- A variety of tests were run for the three
cent in Japan. Chart 1 shows the behavior of Chart 2 groups of countries—the G-3, the G-7, and the
gross debt as a share of GDP for the United Real long-term goverment G-13, the latter being the largest group of
bond rate, 1970—87 countries for which data over the entire sam-
States, Japan, and Germany combined (G-3), (Percentage point)
as well as for the G-3 plus France, Italy, the ple period were available. Different concepts
United Kingdom, and Canada (G-7). In addi- for the level of government, the definition of
tion, 13 OECD countries for which the statisti- the deficit, the use of gross or net debt ratios
cal information was available were aggre- and saving rates, the type of exchange rates
gated to provide the G-13 group. In each of used for aggregation, and various interactions
these groups the weight of a country was among the variables were used in order to
based on its gross domestic product, and the gauge the robustness of the results. The re-
conversion into dollars, on average official ex- sults show significant positive linkages be-
change rates. tween real interest rates on long-term govern-
Chart 2 shows the behavior of real long- ment bonds and the ratio of government debt
term government bond rates, reflecting the to GDP over the 1979-87 period. An increase
average for the same groups of countries. A in the ratio of public debt to GDP by 1 per-
comparison of Charts 1 and 2 reveals some centage point increases real interest rates by
common trends: Until 1974, the ratio of debt about 20 basis points, while an increase in the
to GDP was falling and so was the real bond deficit by a similar magnitude has an effect
rate. Between 1975 and 1980, there was some that is eight to ten times as strong. Higher
increase in the debt ratio, as well as in the real stocks of reserve money lower real interest
Source: IMF Working Paper 91/6.
bond rate. From 1980 to 1985 the share of debt rates, at least in the first year, while higher
in GDP rose very fast, tracking closely with levels of economic activity have the opposite
the real bond rate. After 1985, however, the effect. Higher levels of net private saving were
trends seem to diverge, with the debt ratio seen to reduce interest rates, as expected.
continuing its upward trend, though at a demand for savings (including investment
slower pace, while the real bond rate declined and fiscal deficits) and the total domestic sup-
Changes in government activity
somewhat. ply. For the world as a whole, the demand for Increasing ratios of government debt to
financial savings must be equal to the supply GDP, in addition to raising interest rates, also
Government debt and interest rates of financial savings. International interest have generally harmful consequences for the
We can no longer assume that the fiscal rates must play a major role in bringing level and pattern of government activity. An
deficits of any country will necessarily be about this equilibrium. This means that the increase in interest payments on the public
financed by the financial resources of that impact of fiscal policy on interest rates is debt must result in (1) a higher fiscal deficit,
country alone, or that there must be a strong more meaningful in a global context. This, of (2) a higher level of taxation, or (3) a crowding
correlation between the total domestic course, does not mean that the fiscal policy of out of other public expenditure. Of course, the

Finance & Development I December 1991 31

©International Monetary Fund. Not for Redistribution

 increase in interest payments quire higher interest payments
as a share of GDP, in the ab- Changes in interest payments, debt ratios, and reductions in capital
sence of changes in real inter- spending.
tax levels, and capital spending. 198O—87
est rates, must be preceded by (In percent of SOP)
Thus, in the end, govern-
an increase in public debt. For ments that accumulated debt
industrial countries there has to avoid sudden adjustment in
Change m Change in
been far more discussion interest Change in Change in capital tax rates were forced to raise
among economists of debt fi- payments debt ratios lax ralio sponging
taxes to meet the increasing in-
nancing than of debt servicing. United Stales 1.7 13.7 0.5 -0.2 terest payments. The experi-
In fact, while those writing on Japan 1.3 24.0 4.7 -1.0 ence of the OECD countries
developing countries have of- Germany 0.9 11.3 -0.4 -1.1 shows that the growth of pub-
ten focused on the difficulties France 1.3 10.3 3.1 0.1 lic debt will eventually con-
of servicing these countries' Italy 3.2 33.8 6.0 0.3 tribute to the rise of tax levels.
debt, those writing on indus- United Kingdom -0.4 -4.6 2.2 -0.7
2.9
The pressure to increase taxes
Canada 2.8 23.9 -0.4
trial countries have focused far becomes stronger as interest
Ne! norlands 2.5 29.0 2.2 -1.0
more on the process of debt ac- Australia 1.6 0.7 2.3 -0.9 rates become significantly
cumulation. Sweden 2.5 17.2 7.3 -1.6 greater than the growth rates
The growth of public debt Belgium 4.5 52.5 2.6 -1.8 of the economies, as has oc-
has been defended by some Austria 1.5 20.1 1.2 -1.2 curred in the 1980s. In practi-
economists when it is associ- Denmark 4.4 23.7 6.5 -1.2 cally all countries, the ratio of
ated with wars and major pub- Finland 0.6 6.1 2.9 -0.1 tax revenue to GDP rose over
lic investments, since it pre- Norway 0.9 -1.2 1.2 -0.5 the period. Further, that ratio
vents the increase in tax levels, Ireland1 3.2 48.7 5.9 -0.7
seems to have risen the most
Spain' 3.1 29.7 9.0 —
with their attendant distorting in those countries that experi-
Switzerland -0.3 — 1.2 —
influences, that would be re- enced the largest increase in
Source: OECD
quired to finance the tempo- 'Through 1986. debt and, consequently, in in-
rary increase in public spend- — Indicates absence of dala.
terest payments. The increases
ing. How do these reasons in tax ratios were particularly
apply to the growth of public large in Belgium, Canada,
debt in our sample countries? Denmark, Ireland, Italy, Japan,
There were no such excuses for and Spain, and all countries
large-scale borrowing in the OECD countries interest expenditures. In other words, they try that experienced large rises in debt to GDP
during 1970-87. Thus, fortunately, war fi- to raise the primary surplus. The expendi- ratios and in interest payments to GDP ratios.
nancing cannot be an explanation for the tures most likely to be reduced are those that Although this analysis may be seen as too
growth in public debt in this period, as it had do not have strong support from interest simple, it does provide results that are consis-
been for the growth in American debt in the groups and those whose benefits occur only tent with what one would expect from general
1940s, or for British debt during the past cen- in the future so that their present (especially public choice considerations. Thus, the
tury. The explanation must be found some- political) value is low. Governments typically growth in public debt, with the accompany-
where else. encounter the least resistance if they cut capi- ing increases in interest rates, will in time
What about a bulge in public investment? tal expenditure. Other types of expenditures, lead to increases in tax ratios, as well as to
Is there any evidence of it? Our data indicate such as those for wages and salaries and op- changes in the structure of public expendi-
that, on the contrary, the share of general eration and maintenance, are also likely to be ture, with government investment being pro-
government investment spending in GDP fell squeezed. The table shows the negative rela- gressively squeezed out by rising interest
in practically all countries over the period. tionship between higher debt ratios, which re- payments.
Therefore, the argument that debt financing
would be self-financing, if associated with
major (and productive) capital projects, is not
valid for these countries.
So what has been behind the growth of
government debt ratios? For the countries
and period considered here, the major expla-
nation for rising deficits, and hence the in- Vito Tanzi Mark Lutz
a US citizen, is Director of a US citizen, is an
crease in debt ratios, is rapid increases in gov- Economist with the IMF's
the IMF's Fiscal Affairs
ernment "transfers." The largest of these Department. He is the au- Fiscal Affairs Department.
transfers has been for social security, which thor and editor of numer- He holds a PhD from the
demands greater resources because of aging ous books and articles. University of Maryland
populations, higher levels of eligibility of the and has worked at the
aged for pensions, and higher real pensions Council of Economic
per pensioner. Advisers.
As noted above, when countries face large
fiscal deficits, governments may try to limit
them by raising taxes and by reducing non-

32 Finance & Development / December 1991

©International Monetary Fund. Not for Redistribution