3 The Nexus of Tax Administration and Tax Policy in Jamaica and Guatemala

Richard Bird, and Milka Casanegra de Jantscher
Published Date:
September 1992
  • ShareShare
Show Summary Details
Roy Bahl and Jorge Martinez-Vazquez 

It is almost inevitable that a report on taxation in a developing country, especially if prepared by an outside advisor or a donor, will end by pointing out that the primary problem is one of poor administration. Because low-income countries do not efficiently administer the systems they have in place, they fail to collect the true amount of revenue due, the efficiency objectives of the tax structure are not realized, and both the horizontal and vertical equity intent of the nominal tax structure are compromised.

There seems to be a consensus on why tax administration is so poor in developing countries and on what might be done about it. Most analyses center on three areas:

  • The procedures used are antiquated, and staff are poorly paid and badly trained. This view of the problem has led bilateral donors and international agencies to provide massive amounts of technical assistance in tax administration. There are success stories, but in most countries the scorecard probably still shows a poor administrative effort.1

  • The tax systems in developing countries are too complicated for efficient administration with the modest resources typically available to central government tax departments. Frequently, developing countries have imitated the complex tax structures of developed countries, but have found it much harder to replicate tax administration capabilities. Thus, the proposed solution for developing countries has been policy reform in the direction of simplifying the tax structure.

  • The government is unwilling to enforce the existing system. The approach here is to wait for a crisis to move the government to improve its system, or for outside donors to use coercion (close the loan window) or incentives (open the loan window) to “assist” the process.

If the problem is this well understood, the solutions are viable, and the needed technology is available, then one might ask why tax administration in developing countries has been so slow to improve. We would argue that there are three reasons.

  • The modernization of tax structures, which usually involves following the golden tax rule of broadening the base and lowering the rate, does not always lead to a simplification of the tax structure. In general, the complexities of modern tax structures, which are partly due to the complexities of modern economies, require a higher level of administrative effort. A corollary to this rule is that as developing countries industrialize, and as the level of taxation rises, the incentives for evasion and avoidance go up, and tax administration becomes more difficult.

  • Tax policy is used as an instrument of economic policy, and discretionary tax actions are inevitably “special case” tax preferences or penalties that complicate tax administration. Too often such policies are enacted under the assumption that there is no tax administration constraint.

  • All governments, but particularly governments in developing countries, do not want high levels of enforcement, for economic as well as for political reasons. They are willing to accept a certain amount of evasion and leave open loopholes that permit avoidance. Outside advisors often confuse these deliberate policies with poor administration.

This paper is about the nexus between tax administration and tax policy in developing countries, and draws on the experience with tax reform in Jamaica and Guatemala in evaluating these three propositions. These propositions suggest a need to rethink the traditional remedies for poor tax administration in developing countries. To be sure, more training, better staff conditions, improved procedures, and a less complicated tax code remain sound advice. But, as we enter the 1990s, there are other aspects that need to be considered, including perhaps a need to begin breaking with the conventional wisdom of “go slow and keep it simple.” Modern tax structures are complex, and the complexity grows with economic development. One need not survey too many tax systems to realize this. It may be time to move away from advising governments to shun complexities in their tax structures because of the administrative constraint toward relaxing the constraint so that the administration is better able to absorb the inevitable complexity.

The paper is organized as follows. A brief description of the two tax reform projects is given in Section I; Section II considers the influence of the economic and political setting on the probability of successful policy and administrative reform; and Section III describes the reforms that took place in both countries in the 1980s. These themes are analyzed further in Sections IV, V, and VI. A final section speculates on general lessons for tax policy and administration reform in developing countries.

I. The Jamaican and Guatemalan Tax Projects2

In 1986-87, the Government of Jamaica implemented a major reform of its income and property tax structure. The reform package was a success in that it was popularly accepted, the revenue targets were met, and the package was implemented without any major incident in tax administration. However, it was incomplete in that it did not include a restructuring of the indirect tax system or the payroll taxes. The impetus for policy and administrative reform disappeared after the 1987 reform, but in 1990 the Government moved to reconsider the introduction of measures complementary to those approved in 1986. The groundwork for a major administrative reform was also laid in 1986, and though much progress has been made, that effort is not yet complete.

The tax reform was preceded by a careful evaluation of the existing problems, quantification of the tax system, and simulation of the impact of possible changes. The proposals were subjected to public debate, deliberated upon by a blue-ribbon commission, and reported in the media. The second-round reform, presently under way, is taking this same quantitative approach and is modeling the revenue and tax burden impact of alternative reforms.3

In 1987, the newly elected Government of Guatemala introduced a comprehensive tax reform. In the view of many observers this reform package was riddled with problems. It was unpopular with taxpayers, and it ultimately led to a dramatic decrease in compliance. In contrast to the Jamaican project, the 1987 Guatemalan tax reform was done almost in an ad hoc fashion. In part, this was because the Government could not afford the time to “do it right.” Only a few Guatemalan experts were involved, little international expertise was used in assessing and formulating policy, and there was practically no quantification of revenue and burden effects. Consequently, the Government did not have in hand the necessary background information. The Government made no effort to improve tax administration, and the reforms were designed to accommodate the weaknesses of the existing administration. Rather than public debate, there were a number of presentations to the public through the media and public meetings. The presentation put little emphasis on the characteristics of the tax reform itself and much more on the political and social benefits to be derived from the increased tax collections the reform would provide. In retrospect, 1987 was a missed opportunity for successful tax reform, although it should also be noted that changes to the laws introduced by the Guatemalan Congress at the last minute contributed to diluting the quality of the 1987 reform package.

A new government took office in January 1991, and is considering the need for a more comprehensive reform of the tax system. The present reform project in Guatemala is being carried out jointly by the Ministry of Finance, the Policy Economics Group of KPMG Peat Marwick, and the Policy Research Center of Georgia State University. This effort started in 1989 and it is scheduled to end in early 1992. It is a comprehensive reform program in the same vein as the Jamaican effort.4

II. The Economic and Political Settings

Finding the right relationship between tax policy and tax administration in developing countries begins with choosing the right moment to carry out a reform. The Jamaican and Guatemalan cases show that comprehensive tax reform can take place in a weak economic setting. Reform certainly took place in Jamaica in an economic crisis, and is under consideration in Guatemala in a similar crisis situation.


Jamaica was besieged by severe economic problems over most of the tax reform period. This made it difficult for the Government to keep its attention focused solely on improving the tax structure and its administration. The collapse of the bauxite industry in the early 1980s deprived the country of its major foreign exchange earner. The performance of the tourism industry, the second most important source of foreign exchange, was often weak, and recovery proceeded slowly. Continuous disequilibrium in the balance of payments put pressure on the exchange rate and led to a major devaluation of the Jamaica dollar in November 1983. During this period the debt-service burden was very heavy, averaging over 45 percent of export earnings in the mid-1980s. On the domestic side there was a substantial fiscal imbalance, with the budget deficit reaching about 8 percent of GDP in 1985. With not only high inflation but also high unemployment during the period, and the public sector being the most important employer of unskilled workers, it was difficult for the Government to adopt a policy of expenditure retrenchment.

All of this meant that at the time the reform project began, there was a great deal of pressure to take care of the short-run problem of the fiscal deficit mainly by finding ways to raise more revenue. This environment would not have seemed the most opportune time to focus on structural reform—in particular a revenue-neutral structural reform with an emphasis on reducing allocational inefficiencies induced by the existing tax system.

In other ways, however, the time was exactly right for comprehensive tax reform in Jamaica. Prime Minister Seaga had been elected in 1980 with a mandate to make the economy more responsive to market incentives and to get rid of a complex system of controls and regulations. The task was overwhelming but opportunities for improvement abounded. The foreign trade sector was rife with quotas and licensing, to restrict imports and to compensate for an overvalued Jamaica dollar. An inherited import-substitution growth strategy and a complicated tariff structure were in place, and there were substantial price controls, government ownership of some traditionally private sector activities, and very high marginal income tax rates. The Prime Minister’s economic strategy of replacing government controls with market forces fit well with a structural tax reform program designed to “get the prices right.” Moreover, the Seaga Administration won an overwhelming majority in Parliament in a 1984 election, which enhanced the possibility of eventually passing a tax reform bill. Another stimulus to action came from the external donors—the U.S. Government, the World Bank, and the IMF—all of whom were enthusiastic about Jamaica’s plans for tax reform. Finally, the Jamaican income tax had become so onerous, so obviously unfair, and so out of control that there was substantial public sentiment for a major overhaul.


Guatemala’s economy had performed satisfactorily, and by some accounts rather well, prior to 1980. Traditional exports of coffee, cotton, bananas, and sugar, and industrial production fueled by a policy of import substitution within the Central American Common Market (CACM), were the main engines of growth. During 1960-79, GDP grew at an average rate of 5.5 percent. Prior to the first oil shock in 1973, Guatemala enjoyed a particularly stable price level, with an average inflation rate between 1946 and 1972 of 1.5 percent. Between 1973 and 1979 the average inflation rate rose to 12 percent, but never reached the extremely high levels of other Latin American countries. However, observers also see the 1970s as the period when Guatemala developed most of the problems it faced in the next decade. High rates of effective protection with respect to trade outside the CACM, an overvalued exchange rate after 1972, and preferential credit policies for domestic business (which typically were able to borrow at negative real interest rates) resulted in an industrial sector that despite very low labor costs was not internationally competitive and absorbed little labor input.

In 1980 Guatemala entered a deep recession that lasted over five years. By the end of 1985 GDP per capita had fallen to levels of the late 1960s. The depth of the recession was due to a combination of domestic and international factors. Political instability in the region kept away foreign investors. Internal unrest led to capital flight. The international recession of the early 1980s meant higher interest rates for international loans and a deterioration in the terms of trade caused by large reductions in the prices of traditional exports. But perhaps the most important factor was the collapse of the CACM. Domestic economic policy was characterized during this period by a growing fiscal deficit frequently financed by increases in the money supply, and toward the end by multiple exchange rates. Changes in the tax structure in 1983 and 1985 were largely ineffective in eliminating the deficit.

The Guatemalan setting is similar to that in Jamaica in that the policies of the 1970s were a major contributor to the problems of the 1980s, and because the proposed tax reform would be considered by a new administration with a mandate to develop a new economic policy. In many other ways, however, the setting is quite different in the two countries—the most notable difference being that Guatemala has a small, and Jamaica a large, public sector.

The inauguration of the democratically elected Cerezo Government in January 1986 marked a big turn of events not only in Guatemala’s politics but also in its economic performance. The Cerezo Administration started its mandate with a good economic performance. The average growth rate of real GDP was 3.6 percent. But the economy deteriorated in the last two years of the administration. By the end of 1990 it was obvious that the administration would leave the economy in no better shape than when it was inaugurated in 1986.

What happened between 1986 and 1990 to lead to the fundamental imbalances that occurred in the central government budget and the international accounts? In the external sector, imports grew at a brisk pace after 1986, fueled by continuous growth of private and public spending. However, exports grew at a more moderate pace over this period, because of the sluggish performance of nontraditional exports and deterioration in the terms of trade for traditional exports. Net capital inflows declined from a high in 1986, led by a marked drop in private capital net inflows. Despite repeated devaluations of the quetzal, and a floating exchange rate after 1989, Guatemala’s net reserve position deteriorated. By the end of 1990 the Government had defaulted on most of its international loans, including those from the World Bank and the IMF.

By most accounts, fiscal policy was an even greater problem of the Cerezo Government. To be fair, the Administration inherited the disarray in public finances. It took two policy steps to resolve the problem. It first introduced a short-run program intended to close the fiscal gap. This program included a temporary tax on exports (to be phased out gradually), a temporary tax on international phone calls, and a selective consumption tax. The second, and most important, fiscal move of the Cerezo Administration was the tax reform program of 1987. In the same year, however, the Administration began to emphasize the need to undertake an ambitious domestic investment program in social services, mainly education and health. Because many in the private sector saw this program as creating more bureaucracy and exacerbating the government deficit, support for the Government’s fiscal program was weakened.

Politics aside, the fiscal problem in Guatemala has been that the tax structure is not in step with such government objectives as export-led development. More important, the level of taxes is not high enough to cover the direct expenditures and subsidies desired by the Government. Interest and exchange rate subsidies by the central bank have been an important source of the overall government deficit in recent years. The projected overall government deficit for 1991 is Q 2 billion, or close to 6 percent of GDP.

The financing requirements of the public sector were partially offset in the first years of the Cerezo Administration by foreign aid, mainly budgetary support from the U.S. Government. As the government deficit grew, foreign aid was significantly reduced and the Government resorted to monetization of the deficit. This led to an acceleration of the inflation rate, which by the end of 1990 had reached 40 percent. There was an attempt to introduce new tax legislation in the summer of 1990 in the hope of arresting the decline in tax revenues. The measures included the elimination of a number of incentive laws, new taxes on foreign exchange operations, increases in a number of fees, and the introduction of a tax amnesty. Only a few of these measures were finally approved in Congress, and they were of little consequence.

The new government of Jorge Serrano, which took office in January 1991, faced the need to implement a short-run stabilization plan immediately and to offer a plan to control the fiscal deficit. Without a stabilization plan, approved by the World Bank and the IMF, the options for the new government are few: foreign loans are not available, foreign government donations have dried up, and the ability to borrow from the public and the private banking system is very limited. Increased foreign government grants and loans will depend largely on whether an agreement is reached with the international institutions. As of August 1991, a range of possibilities to reduce the current year’s deficit was under consideration, and proposals for a longer-term tax reform were under preparation.

Comparative Economic and Fiscal Positions

Guatemala is the larger of the two countries, with a greater proportion of its population living in the capital city. Its economy is less open to foreign trade, and its income generation is more heavily concentrated in the primary (agriculture and mining) sector. Jamaica has a significantly higher per capita income. Both countries showed negative economic growth through most of the 1980s.

One of the most noticeable differences between the two countries is the level of taxation: Jamaica raises taxes equivalent to nearly one fourth of GDP, and Guatemala raises an amount equivalent to 5 percent of GDP. Of course, there is no definitive way to establish the appropriate level of taxation for a country, so we cannot jump quickly to the conclusion that Guatemala’s taxes are too low. This is a normative issue to be resolved politically. However, one can ask how the levels of tax effort in Jamaica and Guatemala compare with each other and with those in other countries with similar economic structures and at similar levels of development.5 We applied three different models to estimate tax effort for 66 countries for the mid-1980s (Bahl (1971); Chelliah (1971); and Lotz and Morss (1967)). Based on these indices, Jamaica was a high-tax country, ranking ninth or tenth, while Guatemala was a very low-taxing country, ranking fifty-ninth or sixty-first. For example, if Guatemala had taxed at the average level of the 66 countries in the sample in the mid-1980s, it would have roughly doubled the taxes it actually collected. These results also suggest why Jamaica may have been such a good candidate for a revenue-neutral reform.

The fact that Jamaica exploits its tax base more fully does not lead to the conclusion that Jamaica has a “better” tax administration. Some of the difference is due to a greater willingness of the population to pay taxes, and some is due to the broader scope of services financed by general taxation. Still, there is no escaping the fact that the Jamaican system does collect a great deal more in taxes. Jamaica has invested heavily in its tax administration: an extensive training program, significant progress on computerization of its system, and the establishment of a Revenue Board. The results of stable and good leadership in the Ministry of Finance are not easily separated from other influences, but the Revenue Board has protected the basics of the 1986-87 reform and extended it to indirect taxes. The training program has been steadily improved—a training plan has been adhered to since the mid-1980s—and there has been a continuous improvement in income tax administration. In Guatemala, where there is no such body as the Revenue Board, there have been five different ministers since 1988. There is no training plan, and no consistent, long-term program for improving tax administration.

III. The Tax Reforms


The 1986-87 Jamaican tax reform was meant to be comprehensive, that is, to include a restructuring of all taxes and to improve tax administration. In fact, the structural reform only reached the individual income tax, the company income tax, and the property tax.7 It was also proposed to replace the present indirect tax structure with a value-added tax (VAT). Payroll tax reforms were proposed but not enacted. Both are under consideration in the current reform project.8

Individual Income Tax

Prior to reform, the statutory individual income tax base included all sources of income except bank deposit interest.9 In practice, there was no tax on capital gains, and most self-employed income was outside the tax net. There were two rate structures—depending on whether income was above or below J$7,000. The top marginal rate was 57.5 percent. When payroll taxes were taken into account, the marginal tax rate on an income of J$14,000 was well in excess of 60 percent. There was no standard deduction, but taxpayers could qualify for 16 tax credits. These credits were for different purposes, including personal allowances, stimulation of savings, and even employment of domestics. Because credits were not indexed for inflation, their value had been substantially eroded during the early 1980s. The income tax administration did relatively little monitoring of the credit system.

The base of the tax was further eroded by the practice of permitting employers to grant nontaxable perquisites (“allowances”) to employees. These perquisites were a matter of negotiation between employee and employer (including government ministries) and did not have to be reported to the Income Tax Commissioner. The value of these allowances was estimated at 15 percent of total taxable compensation, but for higher-income workers it was much greater.

Simulation of alternative rate and base structures, with a revenue neutral target in mind and with simplification and neutrality as primary objectives, led to the following reform program:

  • Replacing the present graduated rate structure with a flat rate of 33 ⅓ percent.

  • Replacing the 16 tax credits with a standard deduction of J$8,580 a year.

  • With few exceptions, bringing all nontaxable allowances into the base.

  • Including bank deposit interest (above a low ceiling) in the income tax base.

The Government enacted this program after a Tax Reform Committee representing private business, unions, and other nongovernmental groups spent several months scrutinizing and amending the proposals. The income tax reform became effective at the beginning of 1986 and was almost totally operative by the end of the first quarter.

The Company Income Tax

Prior to reform, companies paid a basic rate of 35 percent plus an “additional tax” of 10 percent, but the additional rate could be offset against a withholding tax on dividends (Wozny (1990)). To complicate matters further, there was separate treatment for agricultural companies, incentive firms, and financial institutions.

This system led to three basic problems. First, the tax was complicated, not easily administered, and was unfair to certain types of firms. Second, it discriminated in favor of debt and against equity finance; in favor of retaining rather than distributing earnings; and against risk taking. The “optimal” dividend distribution rate for a firm was about 27 percent of profits—above this amount “additional” profit tax liability would be due. Moreover, in the eyes of investors, dividends were taxed twice whereas interest received from savings accounts was not taxed at all. Third, the reduction in the top individual income tax rate to 33⅓ percent and the reduction in the U.S. corporate rate brought new pressures to lower the company tax rate.

The Government enacted a comprehensive reform of the company tax in 1987. The tax rate was reduced to 33⅓ percent and the “additional tax” was eliminated. This removed the disincentive to larger dividend distributions, and though the Government did not eliminate the preferential treatment of retained earnings, it did bring interest income into the tax base, thereby removing another disincentive to equity finance. Full loss carryforward was introduced, removing some of the bias against risk-taking, and branch and subsidiary companies were given equal tax treatment. The proposal to exempt dividends from personal tax liability was rejected. The Government instead decided on a separate entity approach whereby company profits and dividends each would be taxed at 33⅓ percent.

Was Jamaica’s Income Tax Reform Successful?

The impact of the flat rate tax reform is now being monitored. The major problem, of course, is to separate the effects of tax changes from everything else that is affecting the Jamaican economy. Even so, there are some indications of success. Perhaps the best indicator is the absence of major public discontent with the income tax reform. The press was not critical, the political opposition did not raise substantial objections (even during the recent national elections), labor seems to be pleased with the relatively high standard deduction and with the fairness of a flat rate, and the business community clearly has benefited from the lower company tax rate. To be sure, there was initial resentment over the taxation of interest income—this led to exemption for small deposits—and there was the expected grousing from special interest groups about the loss of tax preferences. The important point, however, is that the public adjusted to the initial shock of the change; and though taxes are never as low or even as fair as citizens would like, the new system would appear to be much more palatable than the previous one.

Since no monitoring program was put in place for the reform, it is not possible to estimate the effect of the structural changes. The evidence, however, suggests the following:

  • The Jamaican economy grew rapidly in the immediate postreform period. Few would suggest that this growth was principally due to the tax reform, but it could be argued that so favorable a performance of the Jamaican economy could not have taken place under the old tax regime.

  • The results of the 1990-91 project show that the combination of the lower, flat rate, the standard deduction (increased in 1990), and the broadened base did not worsen the regressivity of the system, and that these measures together with the taxation of interest have improved horizontal equity.10

  • The revenue neutrality target was not met: the reformed system produced faster revenue growth than the prereform system would have. This led the Government to increase the standard deduction in 1990.

Administration Reform

As regards the relationship between policy and administration in the Jamaica project, much of the administration work had a life of its own and was unrelated to the policy design. On the other hand, some of the administrative improvements were directly led by the policy design.

The prereform administration was characterized by an acute shortage of skilled staff. DeGraw (1984) reported that in 1983, a time when increased revenue mobilization was at a premium, there were 150 vacancies among the 449 positions authorized for the Income Tax Department. A disproportionately large number of these were technical positions. Throughout the tax administration service, there were too few skilled staff.

The reasons for the staffing problems were the same in Jamaica as in other developing countries. Salaries were too low, even given the job security and prestige that a government post might offer. In 1983, a trained accountant earning J$9,000 in the Income Tax Department could have earned J$14,000 with a private sector accounting firm. There was no formal career development program and little opportunity for promotion. In the case of the Customs and Excise Department, personnel were recruited primarily out of secondary school; entrants had little background in accounting. To compound the problem, no adequate training program was in place in 1983.

The methods used to assess and collect taxes in Jamaica were inadequate. There was no unique numbering system for either businesses or individuals; hence there was no up-to-date master file of taxpayers under the sales or the income tax. The system was completely manual, that is, there was little if any use of the computer other than to print bills. This effectively ruled out the use of third-party information, the cross-checking of sales and income tax returns, and so on. The income tax was essentially a pay-as-you-earn (PAYE) levy, and there was little if any use of presumptive assessments on hard-to-tax groups, such as self-employed professionals. The major problem was record keeping. The income tax file room was inadequate in size and all records were kept manually. Files were regularly misplaced or lost, and records were frequently out of date or incomplete. There was no monitoring of the performance of the tax system. No annual statistical volume reporting the distribution of taxpayers by taxable income brackets was produced, and there was no revenue forecasting model.

The reform project addressed several of these issues. The revenue services have been reorganized to better integrate assessment and collections; an extensive training program for revenue agents has been completed; a number of courses for income tax assessments and collections have been completed; and computerization of the revenue services is well under way.

The reform itself also had an impact on the administrative upgrading. The proposal to replace the existing system of extended excises with a VAT defined a training program for both the revenue agents and the sales tax inspectors. Moreover, the simpler income tax system seems to have made it possible for the Income Tax Department to concentrate more on enforcement of the system. There was a marked increase in the number of employer and individual returns filed, and an increase in the number of new enrollees.11 Audit and special investigation activities in the postreform period, however, do not differ much from the prereform period.

It is difficult to infer much from available data on income tax collection activities, in part because of changes in reporting, in part because income tax administrative activities might be lumpy, as, for example, when campaigns12 are undertaken; and in part because even these statistics may be too aggregated to describe any changes in administrative activities that may have taken place.


As noted above, tax reform in Guatemala has a checkered history. The military regime attempted a stabilization plan in 1983 with the help of the IMF. The plan called for a reduction in public expenditures and a tax reform. The latter was ambitious in scope: a new VAT was introduced, the income tax was modified, the stamp tax was eliminated, and export taxes were reduced. But the 1983 tax reform was preceded by little preparation, especially for such a radically new tax as the VAT. Predictably, the reform was accompanied by considerable confusion among taxpayers, as well as considerable resistance. The VAT rate was reduced from 10 percent to 7 percent only a few months after the introduction of the tax, and stamp taxes were reintroduced in 1984. The attempt by the military regime to introduce new fiscal measures in April 1985 met with strong opposition from the private sector.14

The program that is of most interest to us is the 1987 tax reform. The changes in the tax code affected not only individual and corporate income taxes but also the VAT, the property tax, and other taxes.

Individual Income Tax

The 1987 reform touched all aspects of the tax. Most of the changes, however, had the effect of narrowing the tax base. The increases in exemptions and deductions, particularly personal and dependent deductions, contributed significantly to the erosion of the base. The law also increased deductions for payments of professional fees and education and other family costs and provided an unlimited deduction for paid insurance premiums. The average household filing taxes can deduct nearly Q 20,000, which is over 150 percent of the country’s average household income. The other main factor contributing to the narrowness of the tax base is a long list of exemptions, most prominently the exemption of interest income from financial deposits and most bonds. Repealing this measure alone would almost double personal income tax collections. The law also exempts holiday bonuses (aguinaldos) in amounts up to two months’ salary, pension and retirement income, dividends and, de facto, most capital gains.

The reform also introduced a presumptive income tax for professionals who fail to file a tax return, a potential base-broadening measure. The presumptive (absolute) levels of income, however, were specified in the law, and with inflation, they have already become obsolete. By 1990, most professionals would find it advantageous to pay taxes according to the fixed presumptive income assigned to their profession in the tax law.

The tax rate schedule for individuals was simplified in the 1987 reform by reducing the number of brackets from 68 to 16, and the maximum rate cut from 65 percent to 34 percent. On the other hand, the reform made each rate effective for the entire level of taxable income, thereby creating “notches” at the bottom end of each bracket.

The Company Income Tax

The 1987 reform introduced several important changes in the tax treatment of company income. It reduced the number of tax rates from 5 to 3, and set the maximum rate at 34 percent—equal to the maximum personal rate. But, as in the case of the personal income tax, one rate is applied to all income, creating an incentive to fragment enterprises into smaller units. The reform eliminated the carryforward of losses and required the apportionment of costs between taxable and nontaxable (for example, tax holiday) activities. The reform also introduced a number of enforcement provisions, including limits on deductions (for example, royalties and payments to directors); presumptive measures of income based on revenues for several groups of companies (for example, film distribution and airlines); and minimum taxes for financial institutions (banks and insurance companies), which have to include interest income in their tax bases but are taxed at a lower rate of 18 percent. As a result of these reforms, companies in all sectors of the economy faced an increased liability, but the commercial sector experienced the largest increase.

The reform of 1987 has left a number of business tax issues that will require further attention, such as the tax treatment of foreign source income, the tax treatment of interest income and capital gains and consequently of financial institutions, the integration of personal and company taxes, the distortions in capital investment and finance induced by the absence of an inflation adjustment, and the widespread existence of costly fiscal incentives. The present tax reform project is addressing these issues.


The reform of 1987 maintained the 7 percent rate, but it introduced three major changes to the administration of the VAT. First, current period credits for the tax paid on a firm’s purchases were limited to inputs of goods and services directly used in the process of generating the taxable sales. This measure had the goal of reducing the level of fraudulent credits, but it went too far; for example, under the present law the VAT paid on a typewriter and other office equipment needed for administrative purposes is not deductible from the VAT collected by the firm on its sales. Although the level of fraudulent credits may have been reduced, this has been accomplished at the price of increasing compliance costs for honest taxpayers and probably also administration costs, as well as tax cascading. Second, VAT paid on equipment and other capital inputs could be credited over a five-year period rather than being entirely refunded in the current year. Third, traditional exports and construction were stripped of their zero-rated status and are now treated as final consumers. These two latter reforms were revenue-raising measures, and the last one probably also helped reduce administration costs, though at the expense of introducing economic distortions. With the exception of these three measures, the basic structure of the VAT is sound. The rate of 7 percent, however, is quite low by world standards.

The most vexing problems with the VAT concern administration. Among the administrative problems, perhaps the worst is the slowness of the government refunding of excess credits; frequently, the policy has been one of auditing all (and only) those taxpayers who ask for a refund, no matter how small the refund is. Paying a small amount of tax saves most would-be evaders from government scrutiny.

Property Tax

The property tax underwent major changes in the 1987 reform. The previous two taxes—the national territorial tax and the municipal tax—were unified in a new single tax. Under this tax, the entirety of the taxpayer’s property holdings are added up to a single base and then a progressive rate schedule is applied—although again, as in the case of the income tax, a proportional rather than a marginal rate structure. These reform features have contributed significantly to the unpopularity of the tax, and the implementation of the new tax system has exacerbated this unpopularity. Because tax records were incomplete and assessments outdated, the Government undertook a massive reappraisal of properties based on self-assessments by property owners. Many taxpayers, especially industrial and commercial landowners, boycotted the self-assessment process. The recording and processing of the new information has been very slow, because of the lack of proper planning and because of problems with software and hardware. These administrative problems and the low revenue yield of the property tax (less than 2 percent of total tax collections) have made it, at times, a candidate for elimination from the tax system.

Was Guatemala’s Tax Reform Successful?

This program initially increased the central government share of GDP from 8.9 percent in 1986 to 10.1 percent in 1988, which was the revenue target for the reform. However, the share of government revenues fell back to 9.1 percent in 1989 and plunged to 6.5 percent of GDP in 1990. It is still not clearly understood what has been behind this dismal performance, but a dramatic decrease in compliance seems likely. Simulations with the VAT and personal and company income taxes show a decrease in compliance in 1990 alone of over 25 percent. In 1989, while GDP increased by 15.5 percent in nominal terms, central government revenues rose by only 4 percent.

To what extent has the decline in revenues in recent years been a product of the structural changes of the 1987 tax reform? Has the reformed tax structure yielded a greater or lesser flow of revenue than the pre-1987 tax structure would have yielded? Using simulation models, we found that for the personal income tax, the net effect of the 1987 reform was to reduce projected revenues in 1991 by 37 percent (Q 125 million), compared with what the prereform system would have produced. The main source of this reduction was the increase in personal exemptions granted in 1987. Very few taxpayers—fewer than 1 percent of all potential taxpayers—had an increase in tax liability as a consequence of the 1987 reform. In contrast, the business income tax reform increased projected 1991 revenues by 28 percent (Q 67 million). Most of the increase was due to the higher rates and to disallowing the carryforward of losses. On balance, it appears, however, that the 1987 structural reform may have reduced the built-in revenue elasticity of the tax system.

IV. Proposition No. 1: The Tax Administration Bottleneck Is Not Always Eased by Simplifying the Tax Structure

A view held by tax policy analysts in developing countries (see, for example, Gillis (1989)) is that simplification of the tax structure will lead to a lighter administrative burden. In its most general terms, the argument goes that a simplified tax structure will let administrators shift their attention from the difficult problems of assessment of firms and individuals that have been given special treatment, monitoring of eligibility for exemptions, and classification of income or commodities to more productive activities such as development of a master file of taxpayers, collections, and audit. In this view, complex provisions in tax structures that demand considerable monitoring and sophistication, generally yield poorer outcomes than rougher, less-refined provisions (see McLure and Pardo, Chapter 4 in this volume). The following are suggestive of the kinds of simplification that may lead to such improvement:

  • Replacing Jamaica’s system of over 60 indirect tax rates with a single-rate VAT would allow sales tax collectors to spend less time classifying goods and more time enforcing the system.

  • Broadening the base of Guatemala’s VAT by making exports the only zero-rated transaction in the system would greatly simplify administration.

  • The substitution of a VAT for export rebates would simplify the approach to adjusting for an overvalued exchange rate and would eliminate the administrative task of monitoring the rebate applications.

  • The removal of special tax treatments, such as industrial tax incentives, would reduce the amount of work required of ministry officials to evaluate applications and to participate in an approval process and monitoring.

Such considerations point toward the desirability of a broader-based tax and a less complicated rate structure, if not a single rate. This has long been the advice of most students of tax design, and this seems to be the trend in many developing countries that have recently enacted tax reforms. Clearly there is merit to the broad-base route to simplification of the tax system. For one thing, it may be easier for the administration to handle because of its uniformity, although special rules are often introduced to facilitate administration. It may also be easier for taxpayers to understand, and therefore comply with uniform rules, although here again special rules (for small taxpayers, financial institutions, and so forth) may facilitate compliance. And, finally, with a broader-based system, there are by definition fewer opportunities for tax avoidance. There is also less opportunity to evade by misclassifying income or commodities, or overstating deductions, exemptions, or credits. Another advantage is noted by Gillis (1989, p. 93) in discussing the Indonesian reform of the mid-1980s. “It was expected that simplification would reduce the scope for corruption, since complexities and ambiguities in tax law were used by tax collectors and taxpayers alike to cloak their transgressions.”

The case for simplification of the tax structure through base broadening—as appealing and essentially correct as the argument may be—can be overstated. More precisely, in some instances it has been overtaken by modernization of developing economies. In fact, modernization of tax systems may in many cases increase the complexity of tax rate and base structures, and the cost of efficient administration. This could happen for a number of reasons:

  • Economies become more complicated in the process of development. The financial structure of domestic companies becomes more complex with more suppliers, middlemen, and financial intermediaries; foreign tax considerations arise in more complicated forms; and the share of income earned from capital increases as local capital markets develop. A true broad-based tax in a more complicated economy poses a formidable administrative task for a low-income country with a tradition of taxing only wages in the formal sector, commodity consumption, imports, and large manufacturers.

  • The change from specific to ad valorem bases in developing country sales taxes implies a switch to book audit from physical inspection. This can be a major shock to the tax administration system, and one that is costly, depending upon whether the country has an existing sales tax structure and on the complexity of the VAT structure chosen. With respect to the latter, the number of rates and the extent to which zero rating is used are particularly important (Casanegra (1990); Due (1990a)).

  • A benchmark of economic development is an increased share of consumption in services. Services are difficult to tax, in part because of assessment problems, even in the industrial countries. Services are often provided by smaller firms or self-employed professionals who are typically “hard to tax.” Tax systems with uniform rules extended to the service sector will require a higher administrative effort.

  • As real property wealth grows, it becomes too important a component of income and wealth to be left out of a truly comprehensive tax base. But good property tax assessment, especially when there is a heavy component of business property and owner-occupied residential property, poses difficult assessment problems.

  • Simplification of the tax laws and the introduction of uniform rules in the computation of taxes frequently means the elimination of special provisions that had been introduced in recognition of limitations of the tax administration.

A broad-based tax may well be the right way to go for a developing economy, and simplification is a goal that is to be strived for, but economic development brings with it the need for more complexity in taxation. This, in turn, places more burden on the administrative infrastructure of the country. Some specific examples of this phenomenon, drawn from the experience in Guatemala and Jamaica, illustrate the range of relationships between tax structure simplification and administrative cost.

Jamaica: Individual Income Tax

Perhaps the best case for simplification of the tax structure as a route to improving administrative efficiency is the individual income tax. Jamaica’s reform is a case in point. There were three parts to the simplification of the tax. The first was the replacing of a system of 16 tax credits, and employer discretion to award nontaxable perquisites, with a single standard deduction. This structural change had the (potential) advantage of allowing the Income Tax Department to shift its administrative efforts away from monitoring allowances and credits (to the extent such monitoring was going on) to audit activities. Another advantage was that the number of taxpayers in the system was reduced by 100,000—nearly one third of the total. Finally, the use of a single standard deduction made withholding an easier process for employers, who now were relieved (theoretically) of determining the eligibility of workers for various types of credits.

Second, the adoption of a single rate in place of a progressive rate structure held advantages for the income tax administration. Reduction of the top rate from 57½ percent to 33⅓ percent could reduce the incentive for evasion and avoidance at the top income levels and presumably the efforts required of the Income Tax Department. However, it still leaves a “notch” problem, when a taxpayer moves from a zero to a 33⅓ percent liability position.

Third, interest income was brought into the system, broadening the tax base significantly. The interest tax is very revenue productive and now accounts for over one third of total individual income tax revenues.15 The collections are made via bank withholding, at the full 33⅓ percent rate. Hence, relatively little administrative burden is placed on the Income Tax Department. It is a tax on the banks rather than on individuals. A privacy law prohibits the Government from inspecting the books of the banks, so there was no possibility for an effective tax at the individual level.

The Jamaican individual income tax is a good illustration of the possibilities for administrative improvements through a simplification of the tax structure. But there also is a dark side to this story. The Jamaicans did not take on the difficult administrative problem of taxing the self-employed, who accounted for only 6 percent of total collections in 1986. Nothing in the reform program addressed this problem. By 1990, the percent of total individual income tax collections from the self-employed had not increased. Moreover, all of the nontaxable perquisites could not be eliminated—especially in the areas of housing and automobiles. Some fringe benefits are increasingly difficult to handle because the economy is more complicated (for example, business expenses), and it is necessary to make exceptions to the rule that all compensation shall be taxed at a uniform rate.

Jamaica: Company Tax

A textbook example of how complications in the tax structure impose administrative costs is the Jamaican company tax. Because of the many special features in its rate and base structure, the prereform company tax was not easily administered. The problems were magnified by a shortage of skilled staff and outmoded—and in some cases flawed—operating procedures. Such difficulties of administration not only raise administrative costs but lead to arbitrariness in assessing the tax base and inevitably to some unfairness in the way different firms are treated.

Examples of how a complicated structure can compromise administration relate to capital consumption allowances and inventory valuation. The system of capital allowances is quite complex (and was not changed by the 1987 reform). There are numerous schedules for asset types, special types of allowances for different industries, and incentive laws that provide special treatment to both favored industries and favored types of assets. Income tax officers spend too much time on classification issues at the cost of spending too little time on audit, with the consequence that monitoring is lax and abuses are invited. Compliance costs, in one form or another, are also raised by such a complicated system. Large enterprises make use of accounting firms to assist them in compliance, but smaller enterprises can less easily take advantage of the available options. This introduces an unintended but potentially important nonneutrality into the system.

The other example has to do with valuing inventories. The law requires that inventory be valued at the lower of cost or market value, and most firms use the FIFO (first in, first out) method for determining the cost of their sales. However, some firms have shifted to the LIFO (last in, first out) method, which has neither been sanctioned in the courts nor approved by the Commissioner. Others avail themselves of even more advantageous approaches, such as writing off stocks that are over a certain age and excluding the proceeds of their sales from chargeable income, and are successful because the Income Tax Department lacks an effective audit branch.

The complexity was exacerbated by inflation, which in concert with the present tax structure, drove up real company tax rates, influenced investment choices, and provided additional incentives for avoidance and evasion. The law contained no provisions for inflation adjustments, except for the crude 20 percent to 90 percent approximation on an initial first year allowance. Under inflationary conditions this approach will lead to understated capital consumption (to a differential degree for assets of different lives), and FIFO accounting will understate the cost of goods sold. Both practices cause profits to be overstated and, other things being equal, dampen the rate of investment (Wozny (1990)).

Finally, the availability of three important avenues of tax avoidance—the preferential tax treatment of incentive activities, interest income, and capital gains—encouraged enterprises to undertake tax arbitrage: to engage in transactions whose sole purpose is reduction in tax liability. Among the many avoidance techniques observed in Jamaica are revaluation and sale of assets with lease-back arrangements; revaluation and sale of assets with a distribution of the (nontaxable) proceeds to shareholders; and the leasing of capital equipment by firms receiving preferential tax treatment to affiliated firms that do not receive such treatment.

The point that structural complication burdens administrative cost would seem well made here. But, while removing some of the ambiguities in the company tax law and restructuring it to fit modern norms would help, it would also introduce new challenges to the tax administration such as the need to develop expertise to deal with inflation adjustments, transfer pricing, inventory valuation, and capital consumption schedules.

Jamaica: VAT

The adoption of a VAT has been offered as a solution to many of the problems of the present indirect tax system in Jamaica, and in many ways it is a simplification by comparison with the present system.16 However, it does not follow that administration of indirect taxes in Jamaica will be made easier or less costly.

A VAT was proposed for Jamaica as part of the 1986 reform. At the time of this writing, it is planned to implement it in October 1991.17 It is to be a single rate tax, with a relatively short list of exemptions and zero-rated goods, and will extend through the level of large retailers.

It is proposed to replace the present indirect tax system, which has five components.18

  • Excise duties levied at widely differing specific and ad valorem rates on domestically produced items, but not on beer, spirits, and cigarettes, which are subject to consumption duty.

  • Separate or additional taxes, called consumption duties, levied at various ad valorem and some specific rates on nearly all domestic manufactures and similar imported goods. There are two principal rates, a basic rate of 15 percent and a “luxury” rate of 27.5 percent, plus two reduced rates (2 percent and 5 percent), one high rate (50 percent), and several specific rates.

  • Taxes on luxury products, called retail sales tax, levied at ad valorem rates on the sale at retail or the importation of a limited number of consumer durables, such as automobiles and household appliances. The rates on automobiles range from 13 percent to 57 percent and household appliances are taxed at 10 percent. The value for tax includes customs, stamp, excise, and consumption duties.

  • Selective taxes on services, such as travel (departures), hotel accommodation, entertainment (admissions), and betting, gambling, and lotteries.

  • An import surcharge levied as a stamp duty. Broad classes of goods are excluded, including petroleum, drugs, fertilizer, and school and hospital equipment. The basic Jamaican customs duty includes a Caribbean Community tariff and an external tariff. Most rates are between 5 percent and 60 percent, but many manufacturing firms and government or quasi-government agencies enjoy low rates or exemption on imports of machinery, materials, and other goods used in production.

From the point of view of administration, there are objectionable features in the present indirect tax structure. First, it is complex. The levying of four different internal consumption taxes—with different rates, licensing and return requirements, and considerable overlap of the taxes on various firms and commodities—is a source of substantial confusion. Second, there is an excessive number of rates, especially of the excises. Some of these are so low that they are not worthwhile. Third, there is a continued use of some specific rates, without indexing for value changes, particularly on the high-yield traditional “excise” commodities now subject to consumption duties.

There are two reasons why introduction of the VAT may complicate administration. One is the expansion of the number of firms covered under the tax, and the other is the skills required to effectively assess and collect the tax.

The decision to extend the tax through the level of large retailers involves the registration and control of firms that are not now in the tax net. Moreover, the registration and control process will require more work for all registered firms than is the case for the present consumption duty, excises, or retail sales taxes. Due (1988, p. 212) has noted these difficulties with the implementation of a VAT, and underlined the need for a more modern tax administration system, stating that the VAT is more suitable in countries that have achieved higher levels of development.

The other complication is that a higher level of expertise will be required to operate the VAT, hence a more sophisticated and costly administration will be required. This is the main issue (Cnossen (1990)). Though the Jamaican indirect tax rates are generally stated as ad valorem, much of the assessment is done on a physical basis and makes use of notional estimates of selling price. The “excise man” who does this work is different from the sales tax officer who may be an accountant or an auditor. To ascertain taxable turnover and verify compliance with the law, sales tax staff rely on examinations of books of accounts payable and receivable, with cash and bank statements, but not with physical properties and quantities. A sales tax auditor is an expert in analyzing the flow of funds and in detecting the underreporting of sales, but he is not acquainted with the technicalities of production processes and warehouses.

An important conclusion here is that Jamaica’s indirect tax system cannot be transformed into a more modern VAT system without substantial organizational and administrative changes. Excise staff cannot be molded into sales tax staff without considerable changes in personnel, additional training, and perhaps an upgrading of salary and substantially improved career opportunities.

Casanegra (1990) suggests that the transition to the VAT is markedly easier if a functioning sales tax is already in place. Jamaica may be in a situation where the transition to a VAT will be especially difficult. A major problem with the present system has been shortage of qualified staff. Prior to completion of the revenue agent program, most of the inspectors lacked the type of training necessary for effective auditing. The present inspection program is also burdened by operating procedures that are antiquated in some cases and weak in others. The ratio of the number of inspectors to the number of accounts is in an acceptable range, but the frequent visits to enterprises are not true audits. Due (1990b) reports that there is no system of priorities for inspection or guidelines for the inspectors, no system for them to report their findings, and little supervision. Even in the case of the traditional excises, where administration is relatively more manageable and physical methods of control are used, there is evidence that procedures are inadequate and qualified staff are in short supply. For example, Cnossen (1990) reports that consumption duty supervision of the largest beer factory in Jamaica is exercised by only one junior officer, largely on the basis of the brewing book.

Guatemala: Company Income Tax

Reform of the company tax, even if in the direction of broadening the base, will not necessarily simplify the tax or reduce the required enforcement effort in Guatemala. The present system was formulated with administrative ease in mind. The tax law tends to incorporate in the base activities that offer relatively easy tax handles; and the difficult things to tax, logically, are left out.

At a certain point in the development process, these “difficult things” must be dealt with. The following list is probably not atypical for a country at Guatemala’s stage of development.

The Guatemalan income tax is based on the “territoriality” method of taxation. Foreign source income, both passive and active, earned by resident companies—and individuals—is exempt from income tax, and income accruing to foreigners in Guatemala is taxable. The issue is whether Guatemala should adopt a residence-based method of taxation and impose a tax on the worldwide income (domestic and foreign) earned by Guatemalan residents. Adopting a territorial system of taxation and ignoring foreign source income simplifies tax administration, but it also would narrow the tax base and consequently reduce revenues, if in fact foreign source income can be effectively taxed. A territorial system gives incentives to companies operating in Guatemala to undertake fraudulent operations to convert income earned in Guatemala into exempt foreign income. The territorial system also allows for highly inequitable situations to arise legally. Wealthy Guatemalans residing in Guatemala may pay no income taxes if all their income is, for example, interest income earned from foreign deposits.

Switching to a residence-based system of taxation would in principle eliminate many of these problems, but it also would clearly increase the complexity of tax administration. The worst (and not so unlikely) scenario, given the international experience, is that despite the administrative effort the residence-based system would not produce significantly larger revenues.

At present in Guatemala, most interest income—that from financial deposits and bonds—is tax-exempt, and de facto interest and other costs incurred to earn tax-exempt income are deductible from taxable income. This policy, together with a partial integration of the corporate and personal income taxes (dividends are exempt at the individual level), means that a considerable portion of capital income escapes taxation. Broadening the tax base by making all interest income taxable and adopting a more complete integration system will be very attractive from the policy side, but, again, there is no question that the changes would impose new demands on the tax administration. Identifying and tracking interest income will increase the complexity of audits. Taxing interest income effectively will also require the introduction of a withholding system for this type of income. A withholding scheme for interest income, of course, can become very complicated if the exemption of low incomes is attempted. Administrative considerations typically advise against attempting a more orthodox method of integration.19

Regulated financial institutions, mainly banks and insurance companies, are subject to a separate tax regime in Guatemala.20 Generally, the income tax base includes interest income, but tax rates are lower. An interesting twist is that the administration of taxes on financial institutions is largely implemented by the regulatory agency, the Superintendent of Banks. In a way, one could interpret this arrangement as if the Ministry of Finance had contracted out the tax administration function for this sector.21 The problem is that the regulatory agency and the Ministry of Finance have quite different objectives. For example, from the regulatory agency point of view, soundness may dictate conservative reserve policies, but for the Ministry of Finance this will mean lower tax collections. The issue here is whether the administration of the tax should be transferred back to the Ministry of Finance if a unified income tax were to be introduced. Many would argue that it should. But transferring the administration back to the Ministry of Finance will entail substantial specialized training of auditors in financial accounting and actuarial science. Keeping auditors with these high skills in the public sector would also prove difficult, given the salary limits in the public sector.

The lack of an inflation adjustment, coupled with the deductibility of interest payments and nondeductibility of dividends, imparts a significant bias in the Guatemalan tax system toward debt financing. With high inflation rates, the tax system de facto heavily subsidizes debt-financed investment. These distortions could be eliminated by introducing full inflation indexing.22 Indexing the tax system for inflation may or may not broaden the tax base, but it certainly increases the complexity of the tax laws, the general difficulty of administering the tax system, and compliance costs. Full indexing would require extensive training of both tax administrators and accountants and others in the private sector who work as tax preparers.

The elimination of fiscal incentives is a notable exception to the proposition that base-broadening implies more complex administration. In Guatemala, as in many other countries, fiscal incentives are given to firms for all sorts of reasons (regional development and export promotion are common justifications). These actions narrow the tax base and indirectly decrease tax collections by providing opportunities for tax evasion. The proper administration and control of tax incentives is a very demanding task in terms of resources. Broadening the tax base by eliminating most of the fiscal incentives would not require additional administrative resources and could free up administrative resources.

In Guatemala, it appears that the Government does not at present put much effort into monitoring the fiscal incentives program. Does this imply there is no administrative cost? On the surface the answer would appear to be no. However, the lax administration of the present system probably encourages evasion and narrows the tax base. This, and the tax expenditure associated with the incentives, necessarily causes the Government to raise some other tax to make up the shortfall. The result is that there is increased administrative effort implied somewhere in the system. Elimination of the incentive program would thus free some administrative resources.

V. Proposition No. 2: Administration Is Complicated Because Taxes Are Used to Achieve Multiple Objectives

Economic planners in developing countries regularly violate the neutrality maxim in taxation. Tax policy is used as an instrument of economic policy, even though it may not be the appropriate instrument and even though it may not have a significant effect. The reasons why tax policy is used to shape economic choices and macroeconomic conditions vary from country to country, but the major factors are that monetary policy and direct income policies are weak or ineffective, taxation is a tool that is available to the central government, and discretionary tax changes give the appearance that something is being done.

Not often considered is the fact that the use of taxation as an economic policy instrument has important implications for tax administration, usually in the direction of complicating the tax structure and therefore the administrative task. One possible outcome is that the administration is not up to the job, and the economic objectives are not attained. The other is that the tax administration is revamped to implement the program, but this imposes costs both on the government in terms of the other tax administration tasks that now may go undone, and on the private sector in terms of resources spent for compliance. In both cases, the new provisions may open loopholes or create ambiguities that present opportunities for evasion, avoidance, or corruption.

One reason why this problem occurs is because the designers of the tax policy measures in question are usually not “tax men,” and are hence impatient with objections that the tax administration system cannot carry the changes proposed. Too often, they see this as an administrative detail that can be worked out in the short run, or simply want to ignore the administrative constraint because it causes them to back away from the policy program they most want. The international agencies are sometimes guilty here because they usually are in a hurry to get their programs in place, and there is rarely time to get the tax administration system geared up to do the job. Legislators may be even worse once they fix their sights on a politically popular reform program.

Tax policy is used as part of stabilization programs, to change the allocation of resources to better fit government development objectives, and to achieve social goals. The list of tax policy and quasi-tax policy measures used in Guatemala and Jamaica in the 1980s gives some idea of the variety of these discretionary measures. One could speculate about the extent to which these measures compromise the overall tax administration (see Annex II).

Tax Policy and Stabilization

Stabilization programs became more a regular than an extraordinary occurrence in developing countries in the 1980s, and taxation was usually a centerpiece in these programs. Developing economies are by their very nature exposed to external events because they are often dependent on a very few products to earn their foreign exchange, they trade in commodities that are subject to prices set on world markets, most are energy-poor, and many owe substantial sums to foreign governments and banks. There are numerous ways in which worldwide economic changes can be transferred to the local economy, as both of the countries under study here illustrate. U.S. recessions have affected Jamaica’s tourism, changing world demand for aluminum has affected Jamaica’s bauxite industry, the world price for coffee and bananas has affected Guatemala’s financial position, and an increase in the price of oil brings inflationary pressures on both economies. In the case of small countries such as these even natural disasters (for example, Jamaica’s hurricanes and Guatemala’s earthquakes) can have major consequences for the economy and the budget. The imbalances to be addressed usually include a government budget deficit in addition to or in connection with a foreign exchange problem and inflationary pressures. An increase in taxation is, therefore, a reasonable part of the remedial program.

The conflict between policy and administration arises because stabilization programs are by their very nature emergency programs, and administrative preparation may be given little attention. Consider the situation where the government is under immediate pressure to address a revenue shortfall and introduces a new tax or a major new enhancement to the tax system. If the tax administration is not prepared to absorb the change, the results may be less than expected. In any case, the day-to-day job of the tax administration may be compromised, and effort may be diverted from other activities that are more in the long-term interest of the country.

Almost every tax advisor can cite an example. Guatemala’s 1983 tax reform, which was part of a stabilization package, is a case in point (see Bird (1985)). The main objective of the package was to reduce the fiscal deficit with tax increases and expenditure cuts. In the tax area the most important changes were the introduction of a VAT to replace a turnover sales tax with a 3 percent rate and the elimination of the stamp tax. The VAT was hastily introduced without time to either educate taxpayers or plan its administration. It was placed into operation roughly a month after it had been announced. The new VAT had a rate of 10 percent and used the tax credit or invoice method. At the time, it was believed that the tax credit feature would make the tax largely self-enforcing.

The great expectations for the tax quickly turned into disappointment. Because of strong public protests, the VAT rate was reduced from 10 percent to 7 percent only months after its introduction. Because the revenue yield of the new VAT proved to be substantially less than anticipated, the stamp tax had to be reinstated only months after it had been removed from the tax code. The effectiveness of the new Guatemalan VAT was hurt most by the lack of planning and preparation in tax administration. Requests for refunds from taxpayers for whom credit for tax paid on purchases exceeded tax due on sales were much larger than anticipated. The avalanche of refund requests, many of which were thought to be fraudulent but could not be audited, led to the on-and-off suspension of crediting, and consequently reintroduced cascading effects in indirect taxation and damaged the confidence of the public in the tax system.

Tax Policy and Resource Allocation

In most developing countries, the allocation of resources is not in line with the government’s development plan, and tax policy is frequently used to try and induce a change. There are many examples from the Jamaican and Guatemalan cases:

  • to force additional savings;

  • to increase employment of unskilled workers;

  • to conserve foreign exchange;

  • to discourage luxury consumption;

  • to encourage equity investment in companies; and

  • to encourage risk-taking.

Examples of how these special tax provisions complicate administration may be drawn from the experience in both countries.

Jamaica: Individual Income Tax

Economic planners adjusted Jamaica’s individual income tax in an attempt to stimulate both savings and worker productivity. They did this by providing tax credits for savings and for the employment of household helpers and by providing a preferential tax rate for income earned from overtime employment. Each worker made an original declaration of credit entitlements to the employer; this became a basis for calculating his tax. There was little or no monitoring by the Central Government, and little incentive for the employer to verify the truthfulness of the declaration. The employee was more or less free to claim all credits, or to inflate the number of children for his personal allowance, and so forth. On the other hand, since the amounts of the credit were not indexed, their real value had eroded substantially by the time of the tax reform.

Two questions may be asked. First, did these credits have any significant impact on the level of savings and employment in Jamaica? Second, did the monitoring of these credits impose a significant burden on the tax administration in Jamaica? The estimates of the Jamaican project were that the credits were not used by a broad spectrum of the taxpaying population, and that their elimination would not significantly reduce either savings or the employment of household helpers.

  • Only 4 percent of all filers and only 0.2 percent of all PAYE taxpayers claimed the household helper credit in 1983.

  • The “savings credits” were more widely claimed, particularly that for life insurance. But the amounts allowed were small, equivalent to only about 10 percent of total individual income tax collections in 1983. When the amount of displacement induced from private sector savings is considered, it seems doubtful that elimination of these credits would have much effect on the aggregate amount of savings.

The second program was a preferential tax rate applied to income earned from overtime work. The split between overtime and regular earnings was reported by the employer, but there was again little or no monitoring. The amount of income reported from overtime activities was estimated to be between 5 percent and 7 percent of collections in 1983, but this ratio varied substantially across income classes (Aim, Bahl, and Murray (1990)). For the over J$50,000 income group, overtime income was estimated at between 20 percent and 45 percent of regular income, suggesting that “overtime” was simply a tax loophole.

The evidence seems to indicate that neither subvention—credits or overtime—has a significant impact on savings or work effort or on worker productivity and neither imposed new administration burdens. What would have happened if the Government had attempted to enforce these programs? Each taxpayer’s status in terms of how much he invested in each of the savings programs would have had to be determined. Since the savings programs and the Income Tax Department do not use the same taxpayer numbering system, and since the Income Tax Department does not have a complete master file in any case, the problems would have been monumental. The monitoring of the household helper credit would have been even more difficult, even though only a small number of families claimed this credit. Monitoring the overtime provision would have required a combination of audits and spot visits to establishments. These programs were established without any provision for monitoring, and likely with the presumption that they never would be monitored.

Both the system of tax credits and the preferential tax treatment of overtime income were abolished with the 1986 tax reform.

Guatemalan Employment Credit

The present income tax law in Guatemala offers an employment credit equal to 100 percent (50 percent in Guatemala City) of the social security contributions paid by the employer for new workers earning less than Q 500 a month. This measure was introduced in 1987 to reduce the high unemployment rate in the country and to redress the openly pro-capital bias of the tax laws. From direct observation of tax returns, the number of businesses that have claimed this credit appears to have been very small. The Department of Internal Revenue does practically no monitoring of this or other credits and incentives. The income tax law also offers an education credit of 100 percent of teacher salaries paid by the company or employer under the supervision of the Ministry of Education. As in the case of the employment credit, only a negligible number of firms seem to have used the education credit.

VI. Proposition No. 3: Many Governments in Developing Countries May Not Want to Improve the Enforcement of the Tax System

Tax administration in developing countries may be weak because governments have deliberately chosen not to make it stronger. To be sure, no country has attempted, or even should attempt, total enforcement of its tax system. Ensuring compliance with the tax laws to obtain adequate tax revenues is only one objective of government. Other objectives, such as protecting citizens’ right to privacy or their right to due legal representation and procedures, also have to be weighed indicating how far to go in tax enforcement. Full enforcement probably is not a sensible proposition, even from an exclusive revenue-raising perspective. After a certain level of enforcement is reached, the (administrative) costs of raising revenues any further are likely to exceed any additional taxes collected.

Governments in both developed and developing countries appear to apply a level of enforcement that is less than their capability (that is, they deliberately accept varying degrees of tax evasion and avoidance). A study of the Internal Revenue Service in the United States (Malanga (1986)) reports that for each extra dollar spent in administration, the U.S. Government could get seven additional dollars from audits and twenty-seven dollars from servicing delinquent accounts. If the return on stronger tax enforcement is potentially so high, why do governments refrain from exploiting it even in relatively “politically safe” areas such as delinquent accounts? The attitude of the Guatemalan authorities toward tax enforcement raises the same question. Even at times when there was a pressing need for additional revenues, and well-designed enforcement programs with a high expected payoff were available, the Government was reluctant to implement these programs. Frequently, programs for improving administration were ignored and attention was shifted to other “solutions.”

In fact, this behavior is not necessarily irrational. There are a number of explanations.


Tax revenues raised to reduce the budget deficit or to increase expenditures produce short-run effects that are within the elected life of the politician. The politician must weigh the negative feelings of voters about higher taxes against the positive feelings about higher expenditures. The payoff from improvements in tax administration, on the other hand, is seen only after years of effort. To the extent they are successful, some future politician is likely to reap the benefits. To make matters worse, some of the costs of administrative upgrading are likely to be felt in the very short run. There are inevitably transition problems when a country moves to a new administrative setup, and the “shock” effect can create problems of discontinuities in collections, personnel disgruntlement, and taxpayer anxiety because of the uncertainty involved. This set of timing problems and the normally high discount rate of politicians is one reason why tax administration reform is so often on the back shelf.

A corollary is that since enforcement of the tax laws carries both economic and political costs as well as benefits, the preferred level of enforcement for the politician will involve a certain level of evasion or avoidance. Toma and Toma (1986) argue that tax administrators would be responsive to politicians’ desires because they might be rewarded or punished though budget allocations. Hunter and Nelson (1991) find that the congressional power of a state congressional delegation in the United States (as measured by seniority and committee assignments) is a significant negative predictor of the enforcement activity of the Internal Revenue Service in that state. It should not be surprising to find that such political considerations play a similar and even stronger role in developing countries such as Jamaica and Guatemala. The number of taxpayers, especially large taxpayers, is fewer in developing countries, and they may find it easier to manipulate or control government policies because of less well-developed political institutions.


A second reason for less than maximum enforcement has to do with the unwillingness of many governments to fully eliminate evasion. The level of evasion may be expected to vary with the level of economic development, the tax structure itself, and the idiosyncratic values and behavior of taxpayers. Every advisor working in a developing country has heard the same story: “You must understand that people in this country do not like to pay taxes.” However, it is clear that for any level of economic development and so on, the government can choose widely different levels of enforcement. So, the question becomes, “Why is tax evasion allowed to continue?”23

The theoretical literature on tax evasion, and the practice in industrial countries, would lead one to believe that with the right combination of enforcement and penalties, and an administrable tax structure, evasion could be reduced to much smaller levels. But it is not, and the following are some hypotheses as to why not.

  • As noted above, it is politically very difficult to impose widespread and harsh penalties if evasion is widespread, and if some of the major offenders are the wealthiest citizens and the biggest businesses. Such actions are inconsistent with the desire to stay in office for the longest possible time.

  • Evasion could be eliminated or substantially reduced only if corruption in the tax administration were eliminated or substantially reduced, which would be expensive in monetary and perhaps political terms.

  • Tax administration is costly, and the efficiency of assessment, collection, and audit depends on the amounts spent for this function. It is difficult to convince the public to pay substantially higher taxes so they can be more effectively pursued by tax collectors.

  • The economy may actually be served by evasion. Take the case of evasion under the income tax, which is most rampant for the self-employed. In neither Jamaica nor Guatemala is there any serious attempt to reach this sector of the economy by preparing a comprehensive master roll and introducing vigorous enforcement. While such action might increase revenues, it also might reduce the aggregate amount of savings and investment in the economy, since the self-employed are among the leading entrepreneurs in the economy.

Other Objectives of Tax Administration

A third justification that a government might use for choosing a less efficiently administered tax system is that revenue maximization may not be the only goal. There are a number of other objectives that authorities may be pursuing as part of the tax administration system, and some of these may be in conflict with revenue maximization.

One objective is to reduce compliance costs. A level of enforcement commensurate with reducing tax evasion could imply a substantial increase in compliance costs. Tax statutes (by design) and administrative practice can shift, or try to shift, administrative costs to taxpayers. Withholding systems do this with employers, printing costs may be reduced by distributing fewer instructional materials to taxpayers, and so forth. Also, as noted above, tax systems become more complicated as the economy evolves into a more complex set of market structures and the cost of complying with the tax law increases. Compliance costs may also be higher simply because the government has not taken them into account in its tax design. Moreover, compliance costs are often ignored at the time of writing the law. The law, of course, does not state that there is no need to pay taxes if it becomes very inconvenient or very costly to do so. But the government may be willing to ignore delinquency if compliance costs are high. For example, it was probably never imagined that employers in Jamaica would police carefully the eligibility of employees to claim each of the 16 individual income tax credits available.

Another objective that the government may have in designing its tax administration system is to protect the right to privacy. In Jamaica, for example, the tax on interest income is collected by the banks, but the Government does not have the right to make a blanket examination of individual accounts. Bank secrecy laws raise similar problems in El Salvador and Panama.

Casanegra (1990) has suggested that an important goal of good tax administration is to increase voluntary compliance. However, one does not often find actual tax administration behavior directly aimed at increasing voluntary tax compliance. For example, few tax administration policies are targeted at minimizing the amount of overpayments made by taxpayers. Goode (1981) states that some tax administrators have proposed this as an important goal of tax administration. The reality is that few tax administrations attempt to gain taxpayers’ confidence by reducing overpayments. In Guatemala, for example, taxpayers believe that if they overpay they will never receive the refund. Indeed, the only firms audited under the Guatemalan VAT are those requesting refunds. Even in the United States it was recently disclosed that the Internal Revenue Service had not sent out notices to taxpayers for overpayment even though those overpayments had been identified by tax administrators.

The Attraction of Administrative Discretion

Another reason why governments may not want full enforcement is that some tax policies are popular and acceptable in their nominal form, but entirely unacceptable if fully enforced. In such cases, the government creates loopholes for avoidance or simply ignores evasion—so long as minimal revenue targets are reached. A more negative way to state this is to say that “bad” policy design, for instance, tax laws that are difficult to understand and interpret, may be intentional. Ambiguous norms provide more opportunity for interpretation by politicians and even corrupt administrators. Goode (1981) also makes the interesting point that since resources may be used less efficiently in the public sector, it may be that tax evasion is a safety valve in the economy against “excessive” or poorly designed taxes.

An example of deliberate policy by government to hold back on efficient tax administration in order to offset some of the undesirable effects of the legal tax structure is the following. Before the 1986 reform, the Jamaican individual income tax had a very progressive nominal rate structure, a separate rate for low-income taxpayers, and many tax credits that appeared to benefit the same group. This was politically acceptable rhetoric about how an income tax should look in the Jamaica of the late 1970s and early 1980s. This (nonindexed) structure yielded a built-in income elasticity estimated at 2.48. If the tax had been fully enforced the result would have been substantial automatic increases in the tax burden and in the progressivity of the system. Given Jamaica’s economic and political setting, this clearly would not have been acceptable tax policy.

The solution to this quandary was to allow tax avoidance and to monitor compliance to assure an “acceptable” amount of revenue. The Jamaican system provided relief through permitting employer-determined amounts of nontaxable perquisites. These fringes, which were mostly cash payments not requiring proof of expenditure, were stated as allowances for all manner of expenses, including travel to work, housing, entertainment, and laundry. Allowances were widespread in the public and private sectors, with the result that the actual income elasticity of the individual income tax was less than unity in the period immediately prior to the reform.

The 1986 reform abolished most of these nontaxable perquisites and brought in a flat rate. This reversed the problem. The rate structure now had the look of a “big man’s” tax package, and there was need to add some features that would give the lower-income workers some nominal benefits. The Government provided the loophole by giving the Income Tax Commissioner discretion to determine whether laundry and uniform allowances were “necessary.” Within three years of the reform, about 20 percent of Jamaican income taxpayers were receiving one or both of these two allowances.

Some would argue against the thesis that governments want less than fully efficient tax administration systems, that it is in their interest to efficiently steer assessment and collection efficiency in the direction that best suits their objectives. But if the system is transparent, the government will find itself in a position of having to defend its actions to those who are the losers in the tax game. An opaque system makes government less accountable for the outcomes and permits tax administration discretion on a day-to-day and week-to-week basis.

VII. Summary and Conclusions: What Are the Lessons?

This work in Jamaica and Guatemala seems to offer two messages. The first is the traditional “poor country-weak tax administration” story. The theme is that tax structures are too complicated for the administrative resources available, the administrative staff is poorly trained and badly paid, and administrative procedures are inadequate. The second, perhaps conflicting, message is that governments in developing countries choose some degree of poor tax administration as part of their development strategy. They do this because enforcement complicates other goals they want to achieve (for example, protection of privacy), because administrative improvements do not have a strong political payoff, and because tax evasion may serve as a “safety valve” against “excessive” or poorly designed taxes. Taken together, these two points reinforce what all tax advisors should already know: There is no correct way to do tax reform.

The lessons learned from this work fall into two categories. The first concerns the place of tax administration in the reform process. The second is what we do next to improve the state of tax administration in low-income countries.

Tax Administration and Tax Reform

There is no general agreement about the best way to do tax reform. Should we shock the system with big changes, or proceed incrementally? Can administration reform be separated from policy reform, and what is the right order of the two? Is the technology now in place to permit governments to improve their administrative systems and live within their budget constraints? Under what conditions will governments agree to significant improvements in their tax administration? A way to answer these questions is to distinguish among three different views of the administrative constraints to tax structure reform.

The Incremental View

The first view is that going after a comprehensive tax reform is a mistake because it is always hard to find the right opportunity for it, and the interdependency between the different elements of the reform increases the risk of failure. Proponents of this view favor an incremental approach to tax reform.

The incrementalists consider tax administration as an institution that will change only very slowly. It follows that policy changes must be modest. In this view, tax administration is a reflection of the general level of development of the country and there is little that can be done in the short run. Complex or sophisticated tax schemes should be avoided in favor of simpler, more easily administered taxes. A possible corollary of this position is that efforts and resources used to accelerate the pace of change of tax administration institutions may be wasted. The main determinant of change—the general level of development in the country—is outside the control of the tax authorities. A second possible corollary of this view is that if it were possible to index the quality of tax administration, this measure would be highly correlated with the level of development and there would be relatively little variation in administration among countries with the same level of development.

The Insufficient Resources View

A second view is that it is indeed possible to produce significant improvements in tax administration in a relatively short period of time, say two years, if enough resources are invested and adequate expertise is put in place. This view would seem to hold that major administrative reforms can take place independently of policy changes. This would seem to be the view adopted in Guatemala, at least at times, by international donors, who have supported tax administration reform.

One test of the success of this approach is provided by Ecuador’s tax administration project from 1985 to 1989 (financed by U.S. Agency for International Development). This project had some evaluation of policy, but no explicit reform component (Martinez-Vazquez (1986)). The main objective was to increase tax revenues by improving registration, collection, and audit procedures. No effort was made to change the tax laws, which were cumbersome and antiquated. The Ecuador project could not be described as a success. It started with several successful pilot projects, especially in collections, but fell short of its long-term revenue collection objectives. The problem was partly that the income tax structure was deficient, and many other things went wrong operationally, such as procurement by the donor agency and widespread corruption of public officials. But perhaps the heavy emphasis and dependence on computerization was the single most important factor in the failure of the entire project. The attraction of computers was understandable, given the backwardness of the tax administration system and the high level of corruption. However, making the administrative process dependent on computers (and their maintenance and replacement) in a setting where there frequently were no funds even for photocopying seems in retrospect to have been ill-advised.

A second phase of the Ecuador project was started in 1990, in conjunction with policy reforms and administered by CIAT. At that time, the Government gave serious consideration to the creation of a new “elite” team of tax administrators, mainly auditors, directly under the Minister of Finance and entirely disconnected from the regular tax office. These officers were to be put in charge of the largest taxpayers.

The “Policy-First” View

A third view about successful reform is that tax administration projects without policy reform are almost always doomed to failure. The policy-first view best describes Jamaica’s successful tax reform, and it is the paradigm adopted for the design of the Guatemalan project. An Ecuador-type tax administration project would have failed in Jamaica. No matter how streamlined the Jamaican administration might have become, it still would have faced the task of monitoring the 16 income tax credits, a special tax regime for wages earned from overtime, and the ad hoc granting of nontaxable perquisites by private employers. To have left the job at administrative reform—as advisors from the U.S. Internal Revenue Service had proposed for Jamaica in the early 1980s—would have been folly.


An illustration of the three different views is provided by the issue of personnel and training. One of the most difficult questions of tax policy and administration reform projects is what to do about poorly trained and ill-paid tax auditors, programmers, and other key personnel in tax administration. One strategy is to simply accept the fact; the problem of poorly paid public sector employees goes beyond the scope of any tax reform project, and it is better to adapt policy design to this constraint.

The opposite strategy may be to try to create an elite corps of tax administrators who are very well trained and are paid at least competitively with the private sector, and to go on to design tax policy that is clearly beyond the capabilities of the present administration but with hopes it can be handled by the “new” administration. Jamaica has taken this approach in preparing to implement its VAT. The most recent tax reform in Ecuador considered the introduction of full indexation for inflation in the income tax, requiring considerable new skills in both the tax administration and the private sector. A third strategy could be to target specific jobs for training and try to find ways to retain those individuals by upgrading positions.

Lessons for a Better Nexus Between Policy and Administration

This paper is about why tax administration in developing countries is not better. If the bottlenecks to improvements in tax administration can be identified then a strategy to improve matters can be derived. Based on experience in Jamaica and Guatemala, we offer the following guidelines for an improved nexus between tax policy and tax administration:

(1) If the objective is a “big” tax reform, the proper order of business is first policy and then administration. It makes little sense to invest in procedures to better enforce a tax that is unfair and has undesirable economic effects. The success in Jamaica and Indonesia and the lack of success in Ecuador are cases in point. We can also say that policy without administration is likely to be no more successful.

(2) Broader-based taxes are fairer and introduce fewer distortions to economic choices, and in some cases they simplify administration. In other cases—for instance, the extension of the VAT to lower levels in the chain of distribution—they do not. In fact, as economies modernize it is reasonable to expect that the job of tax administration will become more and more complex, even if a “good” tax structure is in place. To prepare for this complexity, countries should invest heavily in training in tax administration, updating procedures, and finding ways to improve the competitive position of government employees in the tax service.

(3) Single rate taxes offer significant administrative and policy advantages.

(4) Government planners often overuse tax policy to deal with economic problems. There is too little consideration of the ability of the government to administer these programs. Tax administration experts should have at least a “seat at the table” in the formulation of new tax policies. The expected benefits of such programs should be weighed against their administrative costs.

(5) Governments should have long-term plans for the improvement of tax administration systems, just as they have long-term plans in virtually all policy areas. A “capital budget” for this particular form of infrastructure could protect improvements in tax administration against short-sighted politicians who are afraid they cannot claim the benefits of the eventual improvements and who shy away from significant expenditures on enforcement.

(6) Monitoring systems for tax administration should be set up to track changes in the efficiency of taxpayer identification, assessment, collection, and audit. This is especially important in the postreform period, when new policies need protection from those with a vested interest in finding ways not to comply.

(7) Government planners (the leadership in the ministry of finance and the planning ministry) should cease using poor tax administration as a shield for poor design of tax policy.

(8) Invest the necessary resources to keep up-to-date statistics on tax performance. This will both provide information necessary to improve tax administration and will also help in the formulation of tax policy.


Exchange Rates of the Jamaica Dollar and the Guatemalan Quetzal Vis-à-Vis the U.S. Dollar(Annual averages)
Jamaica Dollar-U.S. DollarGuatemalan Quetzal-U.S. Dollar
Sources: International Monetary Fund, Government Finance Statistics Yearbook, Vol. 24 (Washington, 1990), and for those with *, Planning Institute of Jamaica.
Sources: International Monetary Fund, Government Finance Statistics Yearbook, Vol. 24 (Washington, 1990), and for those with *, Planning Institute of Jamaica.

ANNEX II Selected Tax Policy Instruments in Jamaica and Guatemala

Following are some of the tax “interventions” used by the Governments of Guatemala and Jamaica during the 1980s. These were introduced primarily for nonrevenue purposes. A reasonable proposition is that such interventions significantly complicate tax administration.

Individual Income TaxTaxes on the Foreign Trade Sector
Credits for savings
Credits for employment of domesticsImport surcharge
Deductions for professional services and insurance premiumsExport rebates
Export taxes
Nontaxable fringe benefitsFiscal incentives for partial exporters
Preferential rate for overtime earnings
Nontaxable Christmas bonusFree-trade zones
Separate rate schedule for low incomesImport licenses
Tax-free interest incomeImport quotas Exemption of imports Protective tariffs
Company Income Tax
Accelerated depreciation
Tax-free interest income and full interest deductibilityDomestic Indirect Taxes
Tax rebates for dividend distributionsExemptions Zero rating
Limited or no carryforward of losses
Tax credit for issuance of bonus shares
Fiscal incentives


Tulloch-Reid Edwin

In the light of our recent experience in Jamaica, it may be useful to make three general observations before commenting on the specific propositions put forward in the paper.

First, our experience has not only confirmed the nexus between tax administration and tax policy but has demonstrated that a well-conceived tax policy developed from a thorough analysis of the tax system and with clearly defined economic and social objectives is the foundation for implementing sound tax administration measures.

Second, our experience has also taught us that tax administrators need to be constantly reminded that—and here I quote Vito Tanzi in his address at the 1991 CIAT Assembly in Washington, “Tax systems do not just raise revenue but are policy instruments that governments use to pursue their own goals.”

Finally, our experience has again taught us that strategic planning, which has been strongly advocated by CIAT in recent years, is the single most important weapon in the armory of the tax administrator. Here we refer specifically to the methodical identification and consideration of the environmental factors that affect the planning and implementation processes.

It is within this context and in the light of the Jamaican experience that we offer comments on the three propositions put forward in the paper.

Proposition No. 1

The tax administration bottleneck is not always eased by a simplification of the tax structure.

In the case of the income tax, administrative bottlenecks were significantly reduced in Jamaica as a result of the simplification of the tax structure under the tax reform of 1987. For example:

  • The cumbersome system of multiple tax credits (16) was eliminated and replaced by a single standard deduction.

  • A single flat rate of 33 ⅓ percent for individuals as well as companies was introduced. This rate replaced, for individuals, a graduated rate with a marginal top rate of 57½ percent, and, for companies, the basic company rate of 35 percent and the additional company tax of 10 percent.

  • Bank deposit interest income was brought into the income tax base with collections withheld by the banks.

  • Nontaxable allowances were also brought into the tax base, thereby limiting the practice of employers and employees “negotiating” the ratio of taxable and nontaxable income.

These changes drastically reduced the volume of tax computations for both taxpayers and officials and released staff to concentrate on the major problem of widespread tax evasion, particularly among the self-employed.

The proposed general consumption tax, a VAT-type tax, is specifically designed to simplify the system of indirect taxation by introducing a single rate of tax to replace the existing complex tax structure consisting of eight different forms of taxes and a multiplicity of rates. However, introduction of the general consumption tax may not remove administrative bottlenecks because it requires not only a background in more sophisticated accounting and computing skills but the administration of a far more complex tax base that includes zero rating, exemptions, and services.

In order to minimize administrative bottlenecks in the implementation of the general consumption tax scheduled to come into effect in October 1991, the following steps were taken:

  • Revenue agents were specifically recruited and trained to ensure that the required managerial and technical skills would be in place.

  • Computer systems for the general consumption tax were enhanced to accommodate the multiple tax type regime; and

  • A modern office was built for the purpose and appropriately equipped for the convenience of staff and taxpayers.

In the case of the property tax, we have continued under the tax reform with the unimproved value (land value) tax base as the simplest and most cost-effective system of property taxation for a developing country. However, the property tax being a wealth tax, the overriding principle of equity will in time dictate that the land value system be modified to include improvements, particularly where the ratio of “improvements” to “land value” is far greater than the norm.

The land value system, modified to include improvements, is not expected to produce any unusual administrative bottlenecks in light of the experience gained over the years by the valuation authority in determining improved values for purposes other than taxation.

Proposition No. 2

Administration is complicated because taxes are used to achieve multiple objectives.

Ideally, the tax administrator would prefer the political directorate simply to indicate the revenue target and to be left free to identify what he considers to be the appropriate sources of revenue, taking into account not only the prevailing economic and social conditions but available resources in terms of manpower, equipment, and technology. However, as the authors have pointed out, if there is a tendency for monetary and direct income policies to be weak, the political directorate must give constituents the appearance that actions are being taken.

The case of taxes being used to achieve multiple objectives and at the same time giving the appearance that something was being done may be illustrated by the announcement made by the Minister of Finance of Jamaica in his 1991/92 budget presentation that it was necessary to provide tax incentives to encourage productivity by workers in the export and tourist industries. Tips received by hotel workers would therefore be free of tax, as well as payments received by workers under approved productivity schemes.

Some tax officials were of the view that these new measures would complicate administration, be difficult to monitor, and would create loopholes in determining taxable and nontaxable income. But the special tax preferences illustrate the use of taxation as an economic policy instrument.

The authors also contend that “things get out of hand” because the designers of the tax policy measures are not “tax men.” Our experience points to the need for the tax administrator or “tax man,” as a part of his strategic planning, to be more closely involved with the design and development of tax policy with a view to “influencing,” as well as providing viable options for the political directorate. This approach would ensure that in any policy change important implications for tax administration would be fully taken into account.

Proposition No. 3

Many governments in developing countries may not want to improve enforcement of the tax system.

Since the 1987 tax reform, Jamaica has been fortunate to have ministers of finance who were fully committed to enforcement measures. However, it is understandable that the level of commitment of individual members of Parliament to the enforcement process often reflects the views of their constituents or special interest groups.

In practice, the effectiveness of the enforcement program is largely determined by the resources committed to it. Quite often, in Jamaica as well as most developing countries, the revenue agencies are not exempt from the overall budgetary constraints to which the entire public service is subjected. Consequently, enforcement measures are usually given priority by international funding agencies.

Notwithstanding these budgetary constraints, Jamaica has performed well in the area of compliance. Our economic and social survey for 1990 shows that revenue yield was maintained at 32 percent of GDP for that year. As the paper shows, Jamaica ranked high in terms of tax effort.

The extent of support received from the Government for compliance and enforcement activities is demonstrated by the initiatives currently being implemented to improve tax administration in Jamaica.

Recent Initiatives Taken to Improve Tax Administration

Arising from the policy decisions taken by the Government in the tax reform of 1987, the following initiatives have been taken to improve tax administration.

Development of Staff Resources

A specially designed revenue agents scholarship program was introduced under which 115 revenue agents were recruited and trained from 1987 to 1990 to strengthen the assessment and compliance capabilities of the six revenue departments. The program consisted of a one-year full-time course for university and college graduates and included classroom and on-the-job training, continuous assessment, and written examinations.

A special professional category of revenue agents with competitive salaries and a defined career path was established within the public service to accommodate this new cadre of tax officials.

An evening program for training serving officers as junior tax auditors was also started. Financial assistance is provided to cover tuition, books, examination fees, and travel expenses. Forty auditors have participated in this program since its inception in 1989.

The U.S. Agency for International Development, in collaboration with Georgia State University, is sponsoring a public finance program in tax policy development and tax analysis to develop the capability of the Revenue Board to monitor the tax system on a continuous basis by undertaking tax policy analysis and tax burden studies.

Tax Compliance Measures

A comprehensive audit program has been implemented to improve compliance through in-depth audits. Experienced tax accountants from the Revenue Board coordinate audit teams drawn from all revenue agencies in auditing groups of companies and large enterprises in all aspects of taxation including income tax, excise taxes, property tax, and statutory contributions. This approach supersedes the practice by which companies were subjected to a number of unscheduled and uncoordinated audits.

Tax compliance certification is required by all importers, building contractors tendering for government contracts over a specified sum, and businessmen who require foreign exchange to travel abroad. A tax compliance certificate is issued on production of satisfactory evidence that all tax obligations have been discharged or suitable arrangements for payment have been made. Since the introduction of this measure, revenues have improved significantly.

A business enterprise numbering system (BENO) has been implemented as a first step in the development of a unique taxpayer identification system for all businesses and individuals. This is an integral part of the registration of businesses for the new general consumption tax.

New Technology

A computer center for revenue services has been established to develop an integrated tax information system for the entire revenue services and to improve the assessment and collection facilities.

Micro-filming and laboratory processing facilities have also been introduced in the Income Tax and Land Valuation Departments to store and retrieve vital assessment and land ownership records in the respective departments. Preliminary steps have been taken to develop a land sales data bank in the Land Valuation Department. The data bank will facilitate more frequent revision of land values for property tax purposes, thereby improving elasticity in the tax yield. In practice, the values are only revised every ten years, which results in loss of potential revenue and gross anomalies in the tax base.

Voluntary Compliance

A modern Revenue Services Center was established in 1991 as part of a network of centers to be built throughout Jamaica over the next five years. These centers will be linked by a computer system to improve services in the areas of tax collection and taxpayer assistance, information, and education.

Taxpayer participation as part of the education process is encouraged through seminars and forums in which tax officials speak directly with the public and invite questions. This approach has been widely adopted in the general consumption tax program.

The foregoing initiatives are being pursued in accordance with our strategic plan, and the program is adjusted from time to time as economic circumstances dictate.

Jaime F. Pineda S.

We agree with the authors of this paper that inadequate tax administration is a major problem faced by developing countries. This problem is characterized by the use of antiquated administrative procedures, poor training of staff and contractual personnel, and insufficient compensation for public officials and employees, at levels clearly lower than the remuneration for equivalent positions in the private sector.

In Guatemala one must add to these problems a generalized skepticism about public management (“the government is a bad administrator”), which has led to a very low level of tax awareness and compliance. The tax ratio has not increased in the past twenty-five years. In 1965-67, on average, the tax ratio (tax revenue as a percentage of GDP) was 8.2 percent, with a direct tax burden (the ratio of direct taxes to domestic income) of 4.7 percent and an indirect burden (the ratio of indirect taxes to domestic private consumption) of 7.5 percent.1 As shown in the paper, in 1990 the tax ratio dropped to 6.5 percent, despite the tax reform of 1987.

To a large extent, inefficient tax administration is reflected in an inability to monitor taxpayers. With a total of about 30,000 registered businesses, only 200 audits a year have been performed on the average, accounting for less than 1 percent of all taxpayers. Moreover, there is real leniency in administrative and criminal prosecution for tax noncompliance, owing to the imperfections of legal procedures, the inadequacy of executory instruments, and the leniency of clerks in charge of procedures. In recent years, the constitutional elimination of the advance deposit on appeals (solve et repete) has led to a large increase in the number of appeals against additional assessments.

The following figures illustrate the level of tax evasion. In the 1966-67 income tax period, of 24,300 registered taxpayers, 49 percent paid no taxes because they reported a fiscal loss, and the amount of deductible costs and expenses of those who did pay taxes averaged 82 percent of gross income. For the period prior to the 1987 tax reform (fiscal years 1985-86), taxpayers reporting a fiscal loss still accounted for 33 percent, but for sole proprietorships and business corporations that reported taxable income, the average relative weight of deductible costs and expenses represented 95 percent of the reported gross income.

The Tax Plan of Guatemala

Guatemala’s 1987 tax reform was implemented under extremely pressing conditions, with the participation of a limited group of Guatemalan advisors and staff and the help of only one international expert. A political struggle developed in Congress between the representatives of the Ministry of Finance and the deputies of the various political blocs, particularly with the involvement of representatives of the business sector and their attorneys (CACIF). The deputies of the ruling party (DCG) wavered on fundamental decisions, especially with regard to income tax. Improvised modifications were made, with textual errors and the elimination of some paragraphs that made the text of several articles unintelligible. As a result of these and other political and business pressures, Congressional Decrees 90-87 and 95-87 were issued, largely distorting the plan originally sent to Congress. Eliminated from the scope of income tax at that time were interest accrued on savings and time deposits, and on bonds and credit instruments issued or guaranteed by the Government, public entities, financial institutions, banks, and private enterprises, whose issues are classified by the Bank of Guatemala’s Securities Commission as first-class; interest on treasury bonds, notes, or bills issued by the Government, its autonomous institutions, and municipalities; and profits on the sale, transfer, or trading of shares, securities, or credit instruments listed on the stock exchange.

When the tax laws that formed the 1987 tax reform took effect, political campaigning and passive resistance by entrepreneurs began, culminating in a three-day business strike. Then, private sector attorneys began to file appeals for unconstitutionality, eventually totaling 28 appeals (mainly against the income tax, the value-added tax (VAT), the stamp tax, and the single tax on property), of which 6 appeals were successful. These actions partially crippled the tax laws, since unconstitutionality only suspends application of the rule in question and the other provisions of the legal text remain in effect.

At the same time as the tax laws that formed part of the 1987 tax reform were being drafted, a seminar was held on restructuring the organization of the Internal Revenue Office. Participants included all heads of departments, sections, divisions, and administrative units, under the coordination of the General Director and Deputy Directors, with advisory assistance from an expert in procedural simplification and administration (Ing. Luis Melgar) and the logistic support of the Research Department. The seminar concluded by recommending a functional organization. There was, subsequently, great resistance to the changes proposed, due especially to fear of elimination of the “proceedings,” an anachronistic mechanism for issuing routine orders from office to office and even from desk to desk within the same unit, without resolving the matter at hand. The modernization of the administrative infrastructure was delayed.

The unit that monitors large taxpayers’ compliance with filing and payment requirements was not institutionalized nor was its effectiveness increased, although a study showed that 8 percent of taxpayers (approximately 2,400 taxpayers) accounted for 95 percent of the income tax and VAT collections.

The auditing methods were, however, changed. The auditors of the VAT Department were combined with those of the Audit Department. A model report for field audit adjustments was designed, and a new concept of verification was formulated, called selective “integrated auditing.” Under this procedure, no longer would all substantiating documentation be verified; the only items analyzed would be for income, costs, and expenses, which determine the amount of taxable income.

Smaller taxpayers (92 percent of the registered taxpayers), whose income tax and VAT payments represent only 5 percent of collections, would be audited selectively through grouping by economic activity and within the same activity, by comparability factors such as levels of purchases and sales, operating capital, paid-in capital, stock turnover in relation to purchases and sales, and average profitability indices. In these cases, instead of sending the tax auditor to the company, the Office of Desk Auditing would require taxpayers to justify certain revenue and expenses and to explain their irregular behavior with respect to a norm estimated from analysis and behavior of the other taxpayers in the same economic activity.

Unfortunately, there was no support for recruiting 300 additional auditors, though these were requested repeatedly for more than four years. Guatemala continues to be a tax haven, where taxpayers know that they will probably not be audited and, if audited, it is now less burdensome for them to pay attorneys to defend them since the advance deposit on appeals (solve et repete) no longer exists. If any additional assessment results, it is possible that government action will become unenforceable because of the statute of limitations or because a tax amnesty will be enacted.

The Political and Economic Framework

Guatemala is at a crossroads because, in agreement with international funding agencies and donor countries, it has pledged to adopt structural adjustment policies so as to achieve monetary, fiscal, and financial stability with modest economic growth in real terms and an increase in per capita income.

Faced with the political impossibility of undertaking a deeper tax reform that would increase the tax burden on those sectors with the highest income and greatest wealth concentration, the Ministry of Finance has had to resort to offering treasury bonds on the open market in order to attract money in circulation, especially quasi-money (savings and time deposits) institutionalized in the banking system, which early in 1991 surpassed Q 5 billion. The financial cost of this undertaking is very high and means greater pressure on the public finances in the next few years, because the bonds were offered at an interest rate of 33 percent on 360-day investments.

If the current government does not undertake an immediate tax reform that substantially raises revenues, it is likely that the fiscal deficit will reach the unprecedented figure of Q 2 billion, equivalent to 6 percent of GDP.

At the time of these comments (June 1991), the Ministry of Finance has sent one bill to issue economic emergency bonds for Q 700 million (forced saving), which has not been very well received in Congress. If approved, it will most likely be enacted together with a one-time special tax on net wealth, which may not yield the expected revenue. As these comments are being written, the Minister of Finance has announced that Congress has been sent a draft Fiscal Adjustment Law, which is equivalent to a new amnesty on taxes, penalties, and interest.

Economic and Fiscal Positions Compared

The comparison with Jamaica is stimulating, since the most conservative segments of the Guatemalan business sector have used as their argument against tax reform the large nominal increases (from 1985 to 1991) in government revenues and expenditures. However, the deficit in 1985 was 5.9 percent of GDP, while the expected deficit for 1991 will equal about 6 percent of GDP, a difference which in relative terms is insignificant.

In 1988, during President Cerezo’s administration, the tax burden reached its highest level equaling 10.1 percent of GDP, while expenditures remained at about 14 percent of GDP during the five years of that government (1986-90). While Jamaica currently collects over 26 percent of GDP in taxes, Guatemala barely exceeds 7 percent.

Another eloquent comparison relates to tax effort, where the authors’ calculations show that Jamaica ranks as a high-tax country—between ninth and tenth, whereas Guatemala ranks as a low-tax country, between fifty-ninth and sixty-first (out of 66 countries). For Guatemala, the decision to modify the current tax structure cannot be postponed. It has had no fundamental changes in the past 30 years and is characterized by one of the lowest tax burdens in the world and a highly regressive distribution. In fact, indirect taxes on consumption have a weight that fluctuates around 85 percent of tax revenue, while direct taxes on income, real property, and equity transfers represent only 15 percent.

Fiscal Reforms

The authors point out that the 1987 reform of the personal income tax reduced the tax base by increasing the basic exempt allowance, the number of dependents, medical expenses, life insurance premiums, and other deductions for education and transportation. These expanded allowances were granted mainly to favor wage earners, whose only income is from providing their personal services. The Social Security Institute’s statistics showed that, of 835,000 workers and employees registered in 1987, only 1 percent earned salaries over Q 1,000 a month, and a typical family of five (parents and three children) had to earn over Q 1,650 a month in order to have taxable income, since their personal deductions amounted to over Q 21,300 a year. Thus, the plan was to tax only approximately 8,300 employees who earned incomes above Q 1,500 a month, generally business executives and senior public officials. In addition, the tax administration was freed from receiving over 50,000 returns with refund claims.

In regard to the authors’ observations on the provision establishing a system of presumptive income for professionals, we accept that the absolute figures established “have become obsolete” due to the effect of inflation and that professionals may find it advantageous to pay the tax according to that presumptive income. Such income was fixed in absolute figures, with no possibility of indexing or periodic adjustment, in keeping with the constitutional limitation found in Article 239 of the Constitution (legality principle) that gives Congress the exclusive power to establish the tax base, which cannot be delegated according to the private sector lawyers who raised objections to the 1987 tax reform.


With regard to the three major changes analyzed by the authors, we shall explain the reasons for the following VAT reforms:

(1) To grant tax credit exclusively for purchases of manufacturing inputs and merchandise for resale, eliminating the tax credit for indirect administrative and sales expenses, including purchases of furniture and office equipment. This reform was based on the fact that auditing of tax credits showed that taxpayers channeled through their businesses the purchase of luxury goods such as televisions, satellite dishes, and all types of electrical goods and home appliances. They also included expenses such as bills from restaurants and supermarkets, air fares for pleasure trips, and business expenses for executives using credit cards (which are also used as a way of augmenting the salaries of officials and employees).

(2) Tax credits for the VAT paid on local purchases and imports of capital goods (investments in manufacturing equipment and machinery) were deferred. Taxpayers were allowed to deduct them in five yearly installments, on each anniversary of the acquisition, to lower the monetization of the economy derived from cash refunds of excess tax credits. At the time of the 1987 reform, because of the freezing of tax credits decreed by the previous government, which also reduced the VAT rate from 10 percent to 7 percent, net VAT credits of over Q 120 million had accumulated. As a result, before the reform, Congress had authorized the issue of VAT bonds at 9 percent for up to a total of Q 80 million.

(3) Traditional exports and construction were excluded from the zero rate. This was justified because traditional agricultural exports would no longer be subject to the special tax after March 1990 and because the effective income tax paid by producers of such goods did not exceed 5 percent of reported taxable income, which was lower than the effective rate of close to 7 percent on the income of professionals and on wages. In the case of construction, despite the fact that large amounts of VAT were refunded to this activity, it was found that the VAT paid on purchases of construction materials had been charged to the cost of construction. The VAT is currently Guatemala’s main tax, despite the fact that evasion is estimated at 20-30 percent, between noninvoicing and the informal economy.

It is hoped that the financial problems stemming from the Government’s obligation to refund excess tax credits will be solved with the establishment of a system for computerized record keeping of taxpayers’ current accounts, which is expected to be fully operational in the first half of 1992. This will allow the Internal Revenue Office to issue, via the computerized system, a tax credit certificate whose value can be applied to payment of any other tax except import duties; or, if the taxpayer so requests, it will be refunded in cash or with bonds, depending on the prevailing monetary circumstances.

Tax on Property

As the authors state, the single tax on property combined the existing municipal tax in the municipal jurisdiction of Guatemala City and three other municipalities of the Department of Guatemala (Mixco, Villa Nueva, and Chinautla). About 2 million self-assessments were expected; only about 1.2 million were actually received. Recording of the self-assessment data for updating the pre-existing tax registers was flawed. The new collection bills were issued consolidating all the properties of each taxpayer, and collections were frequently duplicated. To date, the deficiencies in controlling and billing this tax have not yet been overcome and, despite the fact that the payment was authorized via the banking system, the taxpayers had to go to the tax offices to clear up billing errors.

Currently, Congress has been presented with a draft Law on Municipal Property Tax that seeks to transfer to the municipalities the administration, control, and collection of this tax at a single proportional rate of 6 percent of the value of each property, revoking the single tax on property of the 1987 reform.

The single tax on property did not yield the expected collection results. In fact, from the Q 18.9 million collected in 1986, collections during 1989-91 have remained at about Q 50 million, except for 1990 when collections fell to Q 39.4 million, achieved with high administrative costs and extreme unpopularity among the taxpayers.

According to the authors, “Not often considered is the fact that the use of taxation as an economic policy instrument has important implications for tax administration.” However, there is no getting around the fact that fiscal and tax policy form part of the national economic policy and must be embodied in specific actions. The Ministries of Finance and Economy, along with the central bank (Monetary Board), make up the financial public sector, which must systematically evaluate the behavior of the principal macroeconomic variables while correcting imbalances or deviations from the policies adapted.

It is undeniable that this complicates fiscal and tax administration and eventually violates tax neutrality, but what is needed is to set verifiable goals and objectives, especially in public finance administration. At the Seminar on Data Processing held in Costa Rica in March 1990 under the auspices of the Central American Institute of Public Administration (ICAP), it was recommended that an office of data processing be established under the presidency of each country, to incorporate into the data base the principal macroeconomic variables, development programs and monetary, banking, credit, fiscal and tax policies, with ongoing monitoring of quantitative goals and objectives as well as achievements in economic stability.


    AlmJamesRoyW. Bahl and MatthewMurray“An Evaluation of the Structure of the Jamaican Individual Income Tax,” Chapter 3 in The Jamaican Tax Reformed. by RoyW. Bahl (Cambridge, Massachusetts: Lincoln Institute of Land Policy1990).

    AlmJamesRoyW. BahlMatthewMurray and BruceRiddle“A Program for Reform,” Chapter 4 in The Jamaican Tax Reformed. by RoyW. Bahl (Cambridge, Massachusetts: Lincoln Institute of Land Policy1990).

    BahlRoyW. “A Regression Approach to Tax Effort and Tax Ratio Analysis,”Staff PapersInternational Monetary Fund (Washington) Vol. 18 (November1971) pp. 570612.

    BahlRoyW. (1990a) “Jamaica: Tax Reform: Evaluation and Lessons,”Bulletin for International Fiscal DocumentationVol. 44No. 1 (1990) pp. 3744.

    BahlRoyW. (1990b) ed. The Jamaican Tax Reform (Cambridge, Massachusetts: Lincoln Institute of Land Policy1990).

    BirdRichardM. “A Preliminary Report on the Guatemalan Tax System” (Syracuse, New York: Metropolitan Studies Program of the Maxwell School, University of SyracuseOctober1985).

    BirdRichardM. (1990a) “An Introductory Overview,” Chapter 14 inThe Jamaican Tax Reformed. by RoyW. Bahl (Cambridge, Massachusetts: Lincoln Institute of Land Policy1990).

    BirdRichard M. (1990b) “The Taxation of Services,” Chapter 22 in The Jamaican Tax Reformed. by Roy W. Bahl (Cambridge, Massachusetts: Lincoln Institute of Land Policy1990).

    Casanegra de JantscherMilka “Administering the VAT,” in Value Added Taxation in Developing Countriesed. by Malcolm GillisCarl S. Shoup and GerardoP. Sicat (Washington: World Bank1990).

    ChelliahRaja J. “Trends in Taxation in Developing Countries,”Staff Papers International Monetary Fund (Washington) Vol. 18 (July1971) pp. 254327.

    CnossenSijbren “The Extended Excise Tax System,” Chapter 20 inThe Jamaican Tax Reformed. by Roy W. Bahl (Cambridge, Massachusetts: Lincoln Institute of Land Policy1990).

    DeGrawSandra L. “Current Administrative Procedures of the Income Tax Department of Jamaica and Some Recommended Changes,” Jamaica Tax Structure Examination Project, Staff Paper No. 4 (Syracuse, New York: Metropolitan Studies Program of the Maxwell School, University of SyracuseFebruary1984).

    DueJohn F.Indirect Taxation in Developing Economies (Baltimore: Johns Hopkins University Press, rev. ed.1988).

    DueJohn F. (1990a) “Some Unresolved Issues in Design and Implementation of Value-Added Taxes,”National Tax JournalVol. 63No. 4 (December1990) pp. 38394.

    DueJohn F. (1990b) “Jamaica’s Indirect Tax Structure,” Chapter 15 inThe Jamaican Tax Reformed. by RoyW. Bahl (Cambridge, Massachusetts: Lincoln Institute of Land Policy1990).

    DueJohn F. (1990c) “A General Consumption Tax,” Chapter 17 inThe Jamaican Tax Reformed. by Roy W. Bahl (Cambridge, Massachusetts: Lincoln Institute of Land Policy1990).

    GillisMalcolm “Comprehensive Tax Reform: The Indonesian Experience, 1981–1988,” in Tax Reform in Developing Countriesed. by MalcolmGillis (Durham, North Carolina: Duke University Press1989).

    GoodeRichard “Some Economic Aspects of Tax Administration,”Staff Papers. International Monetary Fund (Washington) Vol. 28 (June1981) pp. 24974.

    HollandDaniel and James R. Follain “The Property Tax in Jamaica,”Chapter 23 inThe Jamaican Tax Reformed. by Roy W. Bahl (Cambridge, Massachusetts: Lincoln Institute of Land Policy1990).

    HunterWilliamJ. and Michael A. Nelson “The Political Economy of Federal Tax Evasion,” paper delivered at the Public Choice Society meeting in New OrleansMarch1991.

    International Monetary Fund (1990a) Government Finance Statistics Yearbook 1990 (Washington1990).

    International Monetary Fund (1990b) International Financial Statistics Yearbook. 1990 (Washington1990).

    “Jamaica Tax Review: Report No. 1” (unpublished; Atlanta: Policy Research Center, Georgia State University1991).

    LotzJørgen R. and Elliot R. Morss “Measuring ‘Tax Effort’ in Developing Countries,”Staff Papers International Monetary Fund (Washington) Vol. 14 (November1967) pp. 47897.

    MalangaFrank “The Relationship Between IRS Enforcement and Tax Yield,”National Tax JournalVol. 39No. 3 (1986). pp. 33337.

    Martinez-VazquezJorge “An Evaluation of Personal Income Tax in Ecuador,” Staff Paper No. 1, Studies of Fiscal Administration in Ecuador (Syracuse, New York: Metropolitan Studies Program of the Maxwell School, University of Syracuse1986).

    Martinez-VazquezJorge “The Taxation of Financial Institutions in Jamaica,”Chapter 10 inThe Jamaican Tax Reformed. by Roy W. Bahl (Cambridge, Massachusetts: Lincoln Institute of Land Policy1990).

    McLure Charles E. Jr. John MuttiVictor Thurony and George R. ZodrowThe Taxation of Income from Business and Capital in Colombia (Bogotá: Ministerio de Hacienda y Crédito Público1988).

    ShoupCarl S.Public Finance (Chicago: Aldine Publishing1969).

    Toma E.F. and M.Toma “A Congressional Control Model of Treasury Revenue Collection,”Southern Economic JournalVol. 53No. 1 (1986) pp. 14154.

    United Nations (1990a) Demographic Yearbook 1990 (New York1990).

    United Nations (1990b) National Accounts Statistics: Main Aggregates and Detailed Tables (New York1990).

    World BankWorld Debt TablesVol. III (Washington1989).

    World BankWorld Tables1989-90 Edition (Washington1990).

    WoznyJames “The Taxation of Corporate Source Income in Jamaica,”Chapter 8 inThe Jamaican Tax Reformed. by RoyW. Bahl (Cambridge, Massachusetts: Lincoln Institute of Land Policy1990).

We say “probably” because we can point to no analysis that compares and tracks indicators of tax administration efficiency across developing countries.

The 1983-87 Jamaican tax project is summarized in Bahl (1990a) and reported in detail in Bahl (1990b). The 1990-91 Jamaican tax project is under way, but no reports are yet available. The 1989-92 Guatemalan project is also under way, and at this time is reported only in a series of technical working papers produced under the direction of KPMG Peat Marwick and the Policy Research Center at Georgia State University.

The initial tax reform project was carried out by the Revenue Board of the Government of Jamaica and the Metropolitan Studies Program of the Maxwell School of Syracuse University. The project began in mid-1983 and the collaboration ended in September 1987. A second round of the tax reform project is now under way, with the Revenue Board taking the lead and the Policy Research Center at Georgia State University doing the technical and advisory work. This project is scheduled for completion in 1991.

Funding for both the Jamaican and the present Guatemalan projects was provided by the U.S. Agency for International Development, but in neither case was the U.S. Government involved in the actual work.

There is some discussion among the IMF economists who developed these indicators, and others who critique them, about whether the term “tax effort” should be used in describing the indices derived. Some have suggested the less pejorative term “tax performance.” Either way, the index has a numerator, which is the actual level of taxes collected, and a denominator, which is an estimate of the taxable capacity of the country. A country with an index greater than unity is one that taxes at a greater rate than we would expect, given its economic structure and the average practice among developing countries. Higher tax effort is not an unreasonable way to describe this situation, though a problem with the specification of the model is another explanation. In any case, these regression models were never meant to give more than a general idea of the difference in the rate of taxation. In this analysis, the results are robust in showing that Jamaica is a high-taxing country and Guatemala is a low-taxing country.

Monetary amounts are presented here in local currency units (Jamaica dollars). Exchange rates are reported in Annex I.

The property tax reform is not discussed here, but it is considered in detail in Holland and Follain (1990).

The VAT was implemented in October 1991.

The system is described in detail in Alm, Bahl, and Murray (1990) and the reform measures in Alm, Bahl, Murray, and Riddle (1990).

This analysis is reported in “Jamaica Tax Review: Report No. 1” (1991).

Detailed statistics for the prereform and postreform periods are presented in Alm, Bahl, Murray, and Riddle (1990).

Periodically, the income tax administration will mount a “campaign” against a particular sector of the economy where delinquency or nonfiling has got out of line. For example, during the course of this project there were special efforts with respect to informal importers, mini-van operators, and companies in arrears by more than a set amount.

Monetary amounts are presented in quetzales. Exchange rates are reported in Annex I.

A good discussion of the Guatemalan tax reforms of the early 1980s is contained in Bird (1985).

It is difficult to estimate the revenue income elasticity of the interest tax because it is a new tax and the coverage has been broadened in each of the three years since its inception. Clearly, however, the elasticity is well above unity. The income elasticity of the remainder of the individual income tax is also above unity since the reform, but not so markedly as interest income.

Parts of this section draw heavily, with relatively little modification, on Bird (1990a, 1990b); Due (1990b, 1990c); and Cnossen (1990).

See “Jamaica Tax Review: Report No. 1” (1991).

Cnossen (1990) has characterized the Jamaican domestic indirect tax system as “an extended excise system.”

The Jamaican scheme of taxing banks, and allowing banks to exclude the income earned on accounts below a certain level, has not imposed great administrative costs. However, because of secrecy laws, the Government has not done a thorough audit of the effectiveness of the program. Such audits seem less urgent when the revenue yield of a tax policy intervention exceeds rather than falls short of expectations.

This is also the case in Jamaica. See Martinez-Vazquez (1990).

In Jamaica, the banks are largely responsible for assessing and collecting the tax on interest income received by individuals.

Guatemala shares this problem with many other countries. For a comprehensive discussion of inflation adjustment issues, see McLure and others (1988).

For a good, conceptual discussion of the issue, see Shoup (1969, Chapter 17).

Congreso National de Economistas, Contadores Públicos, Auditores y Administra-dores de Empresas. Una política para el desarrollo económico de Guatemala (Administratión de la Re vista Economía, Instituto de Investigaciones Economícas y Sociales, Universidad de San Carlos, Guatemala, 1969).

    Other Resources Citing This Publication