(No citing / First draft)

Debt Relief and Private Sector Development in Africa

BOUSRIH Lobna 1 & HARRABI Sana 2

August 2006

1

PhD student in economics at: « Faculté des sciences économiques et de gestion de Tunis ». [email protected] 2 PhD student in economics at: « Faculté des sciences économiques et de gestion de Tunis ». [email protected]

1

Abstract The purpose of the paper is to empirically investigate the link between debt relief and credit to the private sector in African countries using panel method over the period 1988-2004. The idea of investigating the relationship between debt relief and credit to the private sector came from the fact that domestic debt as shown in Jacob Christensen (2004) is negatively affecting the domestic debt; on the other hand debt relief is expected to alleviate domestic debt and then create space for domestic credit which is mostly constituted of governmental credit that crowds out credit to the private sector. The main results of the paper are: (i) credit to the private sector is significantly positively linked to development of private sector. (ii) Credit to the private sector is significantly positively linked to Debt relief. (iii) finally, in the long term debt relief has positive effect on domestic credit to private sector only in countries with good initial institutional quality.

JEL Classification Numbers: Keywords: Debt relief, credit to Private Sector, Domestic Debt, Africa

2

I. Introduction: Private sector development has been identified as key pillar for economic growth in Africa. Banking intermediation, financing gaps and precisely the credit to private sector is becoming of great concern in African countries. Identifying its determinants has been the major issue of several papers. Fiscal deficit, overhanging debt, high reliance of the government on domestic debt to serve it is external debt and interest payments, and consequently higher interest rates are identified as major obstacles for credit to the private sector. Domestic banking systems in most of the African countries, known in majority for underdeveloped financial markets, if existent, have preference for domestic government securitized debt, crowding out the private sector. Added to the crowding out effect exercised by government reliance on the domestic financial market, growing external debt leads to higher interest rates, higher risk premium and consequently discouraging private sector activity. The initiative of Heavily Indebted Poor Countries (HIPC) initiated to relieve resources of the eligible countries is expected to alleviate governmental pressure on the domestic financial sector and limit interest rates increases. The purpose of this paper is to investigate the link between debt relief and credit to private sector. The second section reviews the literature on debt relief and credit to the private sector, the third section is dedicated to stylized facts, the fourth section presents the methodology used to study the examined relationship and the results, and the fifth section presents the conclusions.

II. Literature review: The link between debt relief and credit to the private sector has not been examined by empirical studies according to our knowledge, nor explicitly developed in literature. But the link between domestic debt and credit to private sector and then between debt relief and the domestic debt made us question the empirical link between debt relief and credit to the private sector in African countries, with special focus on HIPC countries. S. Ibi Ajayi (1997) states that the burden of debt has negative effects on credit to private sector through two channels: the first concerns the resources used to service the debt which crowds out directly credit to private sector, and the second is the debt overhang that leads to

3

higher interest rates and to anticipating higher taxes for its servicing which limits private savings and investment and consequently resources for the private sector. Jakob Christensen (2004) emphasizes the crowding out of the private sector by the domestic debt. According to him “when issuing domestic debt, government tap into domestic savings that would otherwise have been available to the private sector” which is followed by a rise of interest rates if flexible, affecting negatively private sector investment, or by a credit rationing and a crowding out of the private sector if interest rates are controlled. He showed through the study of 27 non CFA Sub-Saharan African countries over the period 1980-2000 that “an expansion in domestic credit of 1 percent relative to broad money causes the ratio of lending to the private sector to broad money to decline by 0.15 percent”. Christensen (2004) stated also that in presence of highly concessional external loans and debt relief (under the HIPC initiative), domestic debt is limited, and “normally” the domestic interest rates are not high giving more space for private sector. Samuel N. Ashong (2001), examining the macroeconomic framework for poverty reduction in Ghana, stated that the burden of the domestic debt is the major obstacle to “private sector-led growth development”. He added that the debt relief granted under the HIPC initiative is supposed to freed up resources and make it available for private sector development mainly if some of it is earmarked for paring down the domestic debt.

III. Stylized facts: The HIPC initiative was launched in 1996 by the World Bank and the IMF to reduce the debt of heavily indebted poor countries to free the natural resources of beneficiary countries. To date, 33 African countries are considered HIPC, 25 have already reached the decision point and benefited from at least 80% of debt relief each. The external debt to GDP ratio of HIPC African countries show that the burden of the debt narrowed during the last years. The average of external debt to GDP ratio decreased from 173 percent in average during the 90s to 144 percent on average between 2000 and 2004. Some countries are laying out of the group with higher debt to GDP ratios: Liberia, Congo Republic, and Guinea Bissau with respectively 933, 527, and 353 percent during the 90s, than drop to 441, 356, and 327 percent from 2000 to 2004. Some others, although have benefited from the HIPC 4

initiative, have seen their external debt to GDP ratio stagnating or even going up from 1980 to 2004. During the 80s, Sao Tome and Principe as an example, stood at the ratio of 132 percent, between 2000 and 2004 this rate reached in average 645 percent. Consequently, the sum of total debt services (TDS) in HIPC countries has known a noticeable decrease since the end of the 80s. TDS in HIPC countries stabilized than below 5 $billion (with disregard to the increase of 1994) during the last years. On the other hand, TDS of non HIPCs kept going up driving TDS in all Africa above 25 $billion (see figure.1).

Figure.1: Total debt service in African group of countries (1985-2004) Total debt service 30.00 25.00

African HIPCs

Non HIPCs

Billions $

Africa

20.00 15.00 10.00 5.00 0.00 2004

2003

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

1985

Source: Author calculations using World Bank Data Base.

The decline of the total debt and the recourse for foreign debt especially at concessional rates should lead to the decrease of domestic interest rates. However, real interest rates in Africa have continued rising since the 80s, period during which most of the African countries started liberalizing their financial markets and specially the interest rates. The interest rates raised surprisingly more in HIPC countries than in non-HIPCs. Interest rates in non-HIPCs stayed around 4 percent in average from 1980 to 2004; while in HIPCs interest rates raised from 1 percent during the 80s to 11 percent between 2000 and 2004 with some countries outstanding. In Malawi the real interest rate was 2 and 4 percent respectively during the 80s and the 90s, and

5

picked up starting 2000 to reach 25 percent in average during the period 2000-2004. Similar case is observed also in Sao Tome and Principe starting from an average rate of (-9) percent during the 80s to be higher than 23 percent between 2000 and 2004. Conversely, the DRC is showing a negative real interest rate of (-24) percent during the period 2000-2004. In effect, interest rate in African countries and especially in HIPCs haven’t stopped rising even after the debt relief initiative and the access of the HIPCs countries to loans at very concessional rates at the international level because domestic financial markets are generally underdeveloped and very shallow which make interest rates very sensitive to any rise in the outstanding stock of debt.

Figure.2: Real Interest Rates in HIPC and Non HIPC African countries Real Interst Rates 20

Non HIPC

HIPC

15

10

5

2004

2003

2002

2001

2000

1999

1998

1997

1996

1995

1994

1993

1992

1991

1990

1989

1988

1987

1986

1985

1984

1983

1982

1981

1980

0

-5

-10

Source: World Bank data base, 2006.

This continual increase is basically explained by the financial liberalization of interest rates to end up the financial repression that had been exercised till earlier the 80s. In addition, in African countries, demand for credits is still much higher than the offer. On the other hand, the declining burden of debt and the decrease of total debt service were expected to create fiscal space in beneficent countries. The government debt being alleviated, its fiscal deficit is expected to narrow, resources in the financial system are under lower pressure and financial resources are freed up to be better used for productive investment, basically for private sector investment.

6

Figure.3: Fiscal deficit to GDP ratio Fiscal deficit to GDP (%)

1 0 1980

1982

1984

1986

1988

1990

1992

1994

1996

1998

2000

2002

2004

-1 -2 -3 -4 -5 -6 -7 -8 -9

Non HIPC

HIPC

Linear (HIPC)

Source: World Bank Data base.

Fiscal deficit in average in African countries knew a continual decrease since the beginning of the 80s. In HIPC countries, although at lower pace than in Non HIPC, the fiscal deficit improved during the last years as highlighted by the trend line in figure.3. In 2004, the fiscal deficit in HIPC countries (on average) is around 6 percent of GDP, while it was almost 7 percent in 1980. The ratio of credit to the private sector to GDP stagnated around 20 percent in average during the 80s, the 90s and between 2000 and 2004 in Africa. This general result hides an important variance between credit to private sector in HIPC and in non-HIPC African countries. In non HIPC countries the credit to private sector to GDP ratio has continued rising steadily since the 80s to reach an evarge of 34 percent between 2000 and 2004. However some non HIPC countries are well behind the tendency of the group with very low credit to private sector to GDP ratio such as Angola and Equatorial Guinea with respectively 4 and 3 percent between 2000 and 2004. The highest ratios are observed in South Africa with 135 percent during the same period, the country with one of the deepest and most developed financial systems in Africa, seconded by Mauritius and Egypt both with a ratio of 60 percent during 2000-04. In HIPCs countries the picture is completely different, showing a declining credit to private sector.

7

In HIPCs the ratio of credit to private sector to GDP decreased during the 90s and continued declining in 2000-2004 from an average of 16 percent during the 80s to 13 and then 11 percent during the 90s and 2000-04. Eritrea has the highest ratio of credit to private sector to GDP of 33 percent in average between 2000 and 2004, the equivalent to non-HIPC country ratio average during the same period. The lowest ratio is observed in the Democratic Republic of Congo with 1 percent in average between 2000 and 2004, seconded by Chad, The Republic of Congo, Guinea, Guinea Bissau, and Sierra Leone with roughly 3 percent during the same period.

Figure.4: Credit to the Private Sector in percentage of GDP (Clair: Non HIPC) Credit o Private sector to GDP (%) in 2004 160.00

140.00

120.00

100.00

80.00

60.00

40.00

20.00

0.00 Zambia Uganda Togo Tanzania Sudan Sierra Leone Senegal Sao Tome and Niger Mozambique Mali Malawi Madagascar Liberia Guinea-Bissau Guinea Ghana Gambia, The Ethiopia Eritrea Cote d'Ivoire Congo, Rep. Congo, Dem. Rep. Comoros Chad C African Rep Cameroon Burundi Burkina Faso Benin Zimbabwe Tunisia Swaziland South Africa Seychelles Nigeria Morocco Mauritius Libya Lesotho Kenya Gabon Equ Guinea Egypt, Arab Rep. Djibouti Cape Verde Botswana Angola Algeria

Source: World Bank Data base.

The private sector development is identified as the key to African development. The positive relationship between private investment in African countries and credit to the private sector is highlighted in table….

8

IV. Empirical Investigation

Our objective is to investigate empirically the effect of debt relief on the development of the private sector through its effect on domestic credit for a sample of 52 African countries for the period spanning from 1980- 2004. The approach uses Panel data with fixed effect: first, to underline the effect of domestic credit on the development of private sector and second, to analyze the role of debt relief on the improvement of domestic credit.

The dependant variable is the ratio of private investment to total investment (PvInves) and the independent variables are the ratio of domestic credit to private sector to GDP (CSPY) and the real interest rate (RIR). The estimated equation is the following one: PvInvesit = α 0 + α 1CSPYit + α 2 RIRit + ε it

(1)

The data source is the World Development Indicator 2006 from the World Bank. In equation (1), we introduce the ratio of credit to private sector to GDP (CSPY) and the real interest rate (RIR) as determinant of private investment development. The output of the regression, in table (1), shows that the ratio of credit to private sector to GDP has the expected positive and significant coefficient at 95% significant level, which corroborates with the recent theoretical and empirical literature presented below. Moreover, the real interest rate introduced in the equation as a determinant of private investment presents a negative and significant coefficient at the 99% level.

In the second equation (2) we investigate the relationship between debt relief and the improvement of domestic credit to private sector in African countries. The estimated equation is:

CSPYit = β 0 + β 1 DRit + β 2 RIRit + β 3 FDit + μ it

9

(2)

Where the dependent variable is the ratio of credit to private sector to GDP (CSPY) and the independent variable are the debt relief (DR) the real interest rate (RIR) and the ratio of fiscal deficit to GDP (FD). The output of the regression (1) in table 2 shows that the coefficient of Debt Relief presents a positive and significant sign at 95% significance level. In regression (2) we introduce as explanatory variable the Debt relief and the real interest rate as indicator of credit to private sector demand. this regression shows a positive and significant sign of the coefficient of real Interest rate (RIR) at 95% significance level……………………….,. Debt relief has preserved its positive sign and statistical significance with respect to the precedent regression. In regression (3) we introduce the ratio of fiscal deficit to GDP (FD) as proxy of…(crowding out effect)……………;that can affect the improvement of credit to private sector. The output of the regression shows that the coefficient of fiscal deficit enters with negative and significance sign at the level of 99% significant level………………… . In regression (4) we introduce all explanatory variables which are the debt relief, the real interest rate and Fiscal deficit as a determinant of domestic credit to private sector. The result of the regression in table (2) shows that debt relief has robust positive and significant effect on domestic credit to private sector. The real interest rate and Fiscal deficit have preserved her positive and negative sign respectively.

Finally, we try to test empirically the long term effect of debt relief on domestic credit to private sector, for that we use the dynamic panel method from the period 1980- 2004. The approach uses GMM-

system estimator, which contain both first-differenced and levels equations as developed by Arellano and Bond (1991), Blundell and Bond (1998, 2000).

The estimated equation is the following one:

CSPY t i − CSPY t i−1 = α 0 + α 1 DR ti + α 2 X ti + +α 3 Polity 0 + α 4 DR ti * Polity 0 + (ε ti + η ti )

10

(3)

Where CSPYt i − CSPYt i−1 is the growth rate of domestic credit to private sector, X ti is a set of explanatory variables, including the real interest rate and Fiscal deficit ratio to GDP. The different regression using GMM system method, carried out in table (3), show that Debt relief preserve in most of estimated equation his positive and significant effect on domestic credit to private sector. In this model the interaction effect between initial institution quality and debt relief is also detected. Significant and positive coefficients for the interactive term are also found

CONCLUSION: Debt relief is creating the expected fiscal space, and thereby has positive effect on credit to the private sector in African countries in the short term. Indirectly debt relief is then encouraging private investment. In the longer term, debt relief has to be associated with good institutional quality to ensure the positive effect seen in the short term. Good governance, transparence, and accountability are essential to make debt relief efficient in realizing its objective. Ensuring institutional quality is thereby, again, key for Africa’s development, on which policy makers and should focus more.

11

Appendix:

Table 1: Effect of Domestic Credit on Private Sector Development

Equation

(1) Variables CSPY

0.0013** (1.90)

Real Interest Rate

-0.0018*** (-2.99)

Observation

836

Adj R2

0.015

Note: The dependant variable is the ratio of private investment to total investment

Equation (1)

(2)

(3)

(4)

0.022** (1.96)

0.029** (2.17)

0.025** (2.17)

0.032*** (2.36)

Variables Debt Forgiveness (log)

Real Interest Rate

0.004** (2.23)

Fiscal Deficit

0.004** (2.07) -0.015*** (-2.66)

-0.015*** (-2.42)

Observation

392

261

392

261

Adj R2

0.0126

0.0375

0.032

0.062

*, **, *** are the 10%, 5% and 1% level of risk respectively T-Student in parenthesis

Table 2: Private Sector and Debt Forgiveness 12

Note: The dependant variable is the ratio of credit to private sector to GDP *, **, *** are the 10%, 5% and 1% level of risk respectively T-Student in parenthesis

Equation (1)

(2)

(3)

(4)

(5)

(6)

0.016*** (2.86)

0.060*** (8.35)

0.061 *** ( 8.15 )

-0.014 (-0.57)

-0.205*** (-5.90)

-0.266*** (-4.66)

0.005*** (3.62)

0.004 *** ( 2.95 )

0.005** (2.09)

-0.001 (-0.29)

-0.017 *** (-5.20 )

-0.029*** (-3.30)

-0.047*** (-3.63)

Variables Debt Forgiveness (log)

Real Interest Rate

Fiscal Deficit

0.97*** (12.36)

Polity 2 (Initial level)

DebtForg*Polity2Intial

0.059** (2.41)

0.185*** (3.91)

Observation

392

261

261

30

38

30

Sargan Test

1608.55***

1426.55***

1292.94***

39.14**

160.30***

80.36**

Table 3: Private Sector, Institution and Debt Forgiveness Note: The dependant variable is the ratio of credit to private sector to GDP *, **, *** are the 10%, 5% and 1% level of risk respectively T-Student in parenthesis

13

REFERENCES

Ajayi S. Ibi (1997) “An analysis of external debt and capital flight in the severely indebted low income countries in sub-Saharan Africa” IMF WP/97/ 68 Arellano, Manuel and Stephen Bond. (1991). “Some Tests of Specification for Panel Data: Monte Carlo Evidence and an Application of Employment Equations.” Review of Economic Studies 58: 277297 Blundell, R. and S. Bond (1998): “Initial Conditions and Moment Restrictions in Dynamic Panel Data Models,” Journal of Econometrics, 87, 115-43. Blundell, R. and S. Bond (2000), “GMM estimation with persistent panel data: An application to production functions”, Econometric Reviews 19(3), 321-340 Christensen Jakob (2004) “Domestic debt markets in Sub-Saharan Africa” IMF WP/04/ 46 Samuel N. Ashong (2001) “Macroeconomic framework for poverty reduction within the context of debt relief: the case of Ghana” Paper presented at WIDER Development Conference on debt relief

14