Bangladesh: Selected Issues

This Selected Issues paper on Bangladesh reviews institutional developments in the foreign exchange market since 2002. In 2002, there have been several aspects of the financial system and exchange market in Bangladesh that posed impediments to a floating exchange rate system. The financial system has been dominated by state-owned commercial banks with assets amounting to about 24 percent of GDP and accounting for some 46 percent of industry net assets. Market interventions have been largely confined to building foreign exchange reserves and to countering rare disorderly market conditions.

Abstract

This Selected Issues paper on Bangladesh reviews institutional developments in the foreign exchange market since 2002. In 2002, there have been several aspects of the financial system and exchange market in Bangladesh that posed impediments to a floating exchange rate system. The financial system has been dominated by state-owned commercial banks with assets amounting to about 24 percent of GDP and accounting for some 46 percent of industry net assets. Market interventions have been largely confined to building foreign exchange reserves and to countering rare disorderly market conditions.

III. Potential Benefits of Tax Reform For Bangladesh1

A. Introduction

1. Bangladesh has one of the lowest tax-to-GDP ratios among Asian countries. Despite recent improvements, tax collection in Bangladesh still stands at around 9 percent of GDP, well below the regional average of more than 12 percent (Table III.1). Studies have shown that the low tax intake is the result of excessive use of tax holidays, basic design flaws in the tax laws, and weak tax administration.2 Cross-country analyses and estimates of revenue foregone from tax incentives in Bangladesh find that there is potential to raise revenues by 2 to 4 percentage points of GDP, in broadly equal measure from direct and indirect taxes (IMF, 2007). This would allow Bangladesh to catch up to the regional average, although generating such revenue would require fundamental tax reforms.

Table III.1.

Bangladesh: Tax Revenue to GDP—Selected Asian Countries

(In percent)

article image
Source: IMF staff reports.

The fiscal year ends on June 30;

Non-oil revenues;

Non-resource revenues.

2. To improve the revenue intake in Bangladesh, a series of tax reforms have been recommended. The main thrust of these recommendations is to broaden the tax base to generate more revenue, while lowering marginal tax rates to reduce distortions. In particular, the following measures have been proposed:

  • Redrafting the value-added tax (VAT) and income tax laws to achieve a much simpler tax system that would broaden the base and reduce reliance on inefficient tax incentives;

  • Reforming the current VAT, which is in reality an excise/turnover tax to move towards a modern invoice-based credit system;

  • Reducing and unifying the multiple corporate income tax rates at 30 percent while limiting the use of tax holidays and other tax relief that erode the base.

3. At the same time, Bangladesh suffers from a shortage of public investment, partly due to lack of resources. At around 4 percent of GDP, Bangladesh has a relatively small annual development budget (ADP), compared with other countries at similar levels of development. Although capacity constraints and governance issues appear to be the primary factors affecting ADP spending, revenue shortfalls prior to FY08 reduced the fiscal space available for higher spending (Table III.2). Tax reforms that generate additional and reliable sources of revenue would increase the ability to finance higher and better quality public investment, with potential positive growth effects.

Table III.2.

Bangladesh: Outturn Compared to Original Budget, FY05–08

(in Billions of Taka)

article image
Source: Ministry of Finance, Monthly Report on Fiscal Position.

4. How would a tax reform that generates resources for public investment and reduces tax distortions impact the Bangladesh economy? This paper tries to answer this question by using a general equilibrium model called the Global Integrated Monetary and Fiscal Model (GIMF) calibrated to Bangladesh’s economy. Simulations show that tax reform can enhance growth in Bangladesh by 1 to 1½ percentage points per annum over the medium term both by raising revenues for productive spending and enhancing the efficiency of the tax system. The revenues are best generated through VAT, rather than personal or corporate income taxes. At the same time, if the rate of return on public investment is low, saving the revenue to reduce the deficit would be more beneficial. In the remaining part of the paper, Section B presents the main features of the GIMF model and explains how it was calibrated for Bangladesh. Section C reports the simulation results and Section D concludes.

B. Model, Data, and Calibration

5. The GIMF model is a general equilibrium model that can be used for analyzing tax policy changes. It is intended to help policy makers consider the implications of alternative fiscal (and monetary) policies on macroeconomic variables. The main features of the model are explained in Kumhof and Laxton (2007). In particular, the model allows for (i) liquidity-constrained agents, who cannot smooth their consumption one for one with changes in disposable income; (ii) private investment that responds inversely to changes in the capital tax rate; and (iii) public investment that enhances the productivity of private capital and potentially crowds-in private investment.

6. The model is calibrated using two countries, Bangladesh and the rest of the world. To the extent possible, the parameters for Bangladesh are based on historical values and staff projections. Otherwise, they reflect estimates based on other countries with similar levels of development. The rest of the world is proxied by the U.S., for which data are provided by Kumhof and Laxton (2007).

7. The following assumptions were used to tailor the GIMF model to the Bangladesh economy (Table III.3).

Table III.3.

Bangladesh: Main Assumptions for the GIMF Model

article image
  • Country size. With a population of 150 million, Bangladesh represents around 2.5 percent of the world population and 0.3 percent of world GDP (on a PPP basis).

  • Growth, inflation, and interest rates. Baseline growth is set at 6 percent, in line with the average over the last five years. This implies that output per worker grows at 4 percent, based on a population growth of 2 percent. Inflation is set at 5 percent. The real interest rate is set at 2 percent, based on the historical weighted average cost of borrowing.3

  • Debt and deficit. The debt-to-GDP ratio is set at 40 percent while the overall fiscal deficit is set at 4 percent of GDP, both close to current levels.

  • The share of labor income in GDP is set at 80 percent, based on an estimate by Caselli and Feyrer (2007) for India.

  • Share of liquidity-constrained consumers is assumed to be two-thirds of the population. It is set higher than the typical estimates for emerging market economies (50 percent) and the United States (33 percent).

  • Debt risk premium is assumed to be 4 basis points. This implies that a 1 percentage point increase in the debt-to-GDP ratio leads to a 4 basis points increase in the real interest rate. Given Bangladesh’s low level of financial development and limited integration with capital markets, this risk premium is on the low side of estimates for other developing and emerging countries, which range from 4 to 8 basis points (Rowland and Torres, 2004).

  • Rate of return on public investment is set at 13 percent (baseline assumption). This is based on the standard GIMF assumption, drawn from Ligthart and Suarez (2005), that the elasticity of output with respect to public investment is 0.14.4 Given the difficulty of estimating this value for Bangladesh, we simulate the model also with a higher (17 percent) and lower (10 percent) bound estimate. The deprecation rate of capital is assumed to be 10 percent, both for public and private capital.

C. Simulation Results

8. This section reports the results of four simulations where the tax-to-GDP ratio is raised permanently by 1 percentage point. In the first three simulations, public investment is raised in tandem with tax revenue. In the last simulation, the additional revenue is saved to reduce the deficit. In each case, the results are reported in terms of deviations from the steady state.5

Scenario 1: Raising Tax Revenue and Public Investment

9. Theory predicts that raising taxes would reduce growth, at least in the short run, while raising public investment would raise it. Raising taxes would hurt growth because it would depress aggregate demand. At the same time, public investment can raise the productivity of private capital and even crowd in private investment. The net impact would depend on the particulars of the economy.

10. Simulations show that raising taxes and public investment in tandem would likely lead to a significant and positive growth impact in Bangladesh. The impact is front loaded, with about half of the gains taking place in the first five years. In particular, raising taxes and public investment by 1 percentage point of GDP would boost growth by ¼ to ½ percentage points per annum over the next five years, depending on the rate of return on public investment. Over the long run, output would be higher by 2 to 6 percent in real terms, while private investment would be higher by up to 3 percent. Public investment would crowd-in private investment as long as it has a high or medium rate of return.

Figure III.1.
Figure III.1.

Bangladesh: Main Simulation Results for Scenario 1

Citation: IMF Staff Country Reports 2008, 335; 10.5089/9781451804270.002.A003

Scenario 2: Raising Tax Revenue and Public Investment (and Improving Tax Efficiency)

11. This scenario explores the impact of a tax reform that generates efficiency gains, in addition to raising revenues. The efficiency gains can come from various sources. First, reducing the highest marginal tax rates, while broadening the base, could bring efficiency gains.6 Second, reforms that improve the composition of taxes, in particular by reducing payroll and corporate taxes and increasing reliance on consumption taxes (VAT) would improve production and savings incentives. Third, and more specific to Bangladesh, moving away from highly distortive excise/turnover taxes to a truly modern VAT system could improve efficiency. Finally, the simplification of tax laws and tax administration could reduce the cost of compliance for businesses and improve efficiency. We model these efficiency gains as a one-time increase in total factor productivity (TFP) at the time of reform (by 5 percent).

12. Simulations show that the growth impact would be twice as much as the first scenario, if efficiency gains are also considered (Figure III.2). The impact would be front loaded again, with about half of the gains taking place in the first five years. In particular, growth would be higher by 1 to 1½ percentage points per annum over the next five years. Over the long run, output would be higher by 5 to 10 percent in real terms, while private investment would be higher by 1 to 4 percent.

Figure III.2.
Figure III.2.

Bangladesh: Main Simulation Results for Scenario 2

Citation: IMF Staff Country Reports 2008, 335; 10.5089/9781451804270.002.A003

Scenario 3: Raising Tax Revenue and Public Investment (with Different Tax Instruments)

13. This scenario simulates the impact of raising taxes through three different tax instruments (Figure III.3). The simulation shows that revenues should be generated through consumption tax (VAT), rather than capital and labor taxes. This is because labor and capital taxes create distortions in employment and investment decisions, while consumption taxes are neutral in these decisions. As a result, the highest gain in output is achieved through the consumption tax, while capital and labor taxes yield much lower benefits because they deter private investment and employment, respectively.

Figure III.3.
Figure III.3.

Bangladesh: Main Simulation Results for Scenario 3

Citation: IMF Staff Country Reports 2008, 335; 10.5089/9781451804270.002.A003

Scenario 4: Deficit Reduction Versus Current Spending

14. This scenario shows that a reduction in the deficit financed by higher tax revenues (rather than an increase in public investment) would also have a positive impact, whereas an increase in current spending would not (Figure III.4). In particular, raising taxes by 1 percentage points of GDP, while keeping spending constant, would reduce the deficit by the same amount and public debt by 10 percentage points over the long term. This would lower the real interest rate by 40 basis points, stimulating private investment and growth. In fact, the gains from deficit reduction could be even higher than public investment if the productivity of public investment is sufficiently low (see Table III.4). On the other hand, current spending is strictly inferior to both options, because it neither adds to public capital nor lowers interest rates.

Figure III.4.
Figure III.4.

Bangladesh: Main Simulation Results for Scenario 4

Citation: IMF Staff Country Reports 2008, 335; 10.5089/9781451804270.002.A003

Table III.4.

Bangladesh: Summary of Results: Impact of Tax Reform on Output

(In Percent)

article image

15. Table III.4 summarizes the results of the simulations. In general, tax reform generates positive and significant output gains for Bangladesh. Although raising taxes is usually bad for growth, at least in the short term, simulations shows that raising them in tandem with public investment can generate the opposite effect for Bangladesh. A tax reform that raises revenues and public investment by 1 percentage points of GDP would raise real output by 2 to 6 percent without assuming efficiency gains from tax reform (Scenario 1) or 5 to 10 percent with efficiency gains (Scenario 2). Consumption taxes present the best option for raising revenue, as opposed to capital and labor taxes (Scenario 3). However, if the rate of return on public investment is low, (less than 10 percent), it would be better to save the additional revenue to reduce the deficit (Scenario 4).

D. Conclusion

16. Fundamentally reforming the tax system remains one of the key policy challenges in Bangladesh. This paper estimates the potential benefits of tax reform for Bangladesh by simulating various reform scenarios in a general equilibrium model calibrated to Bangladesh’s economy—the first of its kind to the best of our knowledge. Simulations show that tax reform can enhance growth in Bangladesh by 1 to 1½ percentage points per annum over the medium term, both by raising revenues for productive spending and enhancing the efficiency of the tax system. The revenues are best generated through VAT, rather than personal or corporate income taxes. This suggests that VAT should become the workhorse for revenue generation in Bangladesh, while reforming the income taxes could do wonders in terms of generating efficiency gains. At the same time, if the rate of return on public investment is less than 10 percent, saving the revenue to reduce the deficit would be more beneficial.

Annex

Annex III.1.

Bangladesh: Summary of the Tax System as of July 2008

article image
article image
article image
article image
article image
article image
article image
article image
article image

References

  • Caselli, F. and J. Feyrer (2007). The Marginal Product of Capital. Quarterly Journal of Economics 122 (2).

  • International Monetary Fund, 2007, Bangladesh: Selected Issues, IMF Staff Country Report No. 07/230 (Washington).

  • Kumhof, Michael, and Douglas Laxton, 2007, “A Party Without a Hangover? On the Effects of U.S. Government Deficits,” IMF Working Paper No. 07/202 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Ligthart, J.E. and Suarez, R.M.M. (2005), “The Productivity of Public Capital: A Meta Analysis”, Working Paper, Tilburg University.

  • Poirson, Helene, 2006, “The Tax System in India: Could Reform Spur Growth?” IMF Working Paper, No. 06/93 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Rowland Peter and Jose Luis Torres (2004). Determinants of Spread and Creditworthiness for Emerging Market Sovereign Debt: A Panel Data Study. Borradores de Economia 295, Banco de la Republica de Colombia.

    • Search Google Scholar
    • Export Citation
1

Prepared by Serkan Arslanalp (FAD).

2

A summary of the current tax system is attached as an annex to this paper.

3

The cost of external borrowing is estimated to be 1½ percent in nominal terms, or -½ percent in real terms based on a U.S. inflation rate of 2 percent. The cost of domestic borrowing is 4½ percent in real terms.

4

The elasticity indicates that a one percent increase in public investment leads to a 0.14 percentage point increase in long-run output. The marginal productivity of public investment would then be 3.5, calculated as the elasticity (0.14) divided by the public investment-to-GDP ratio in Bangladesh (4 percent). It is interpreted as the payoff in dollar terms from a dollar increase in the public investment budget. This would yield a rate of return of 13 percent based on the assumption that the payoff arrives over 10 years.

5

Full simulation results are available from the author.

6

It is a standard proposition in public finance that the deadweight loss of a tax (the Harberger triangle) rises approximately with the square of the tax rate.

Bangladesh: Selected Issues
Author: International Monetary Fund