19 Colombia

Teresa Ter-Minassian
Published Date:
September 1997
  • ShareShare
Show Summary Details
Ehtisham Ahmad and Katherine Baer

This chapter provides an overview of the major issues in the area of fiscal decentralization and local government finances in Colombia, an evaluation of the current systems, and a summary of options for reform.

Colombia’s local government finance system has undergone fundamental changes in the recent past aimed at decentralizing the public sector, increasing the autonomy of local and regional governments, and improving the quality of public services provided to the local population.1 Although there has been a gradual shift toward a decentralization of revenue and expenditure responsibilities in Colombia for many years,2 the 1991 Constitution provided a major impulse to the process of decentralization.3 In addition, the popular election of municipal mayors and departmental governors has introduced a dynamic political constituency in favor of decentralization.

At the same time, reform of local public finances has acquired increased importance as the regional and local governments have begun to contribute to macroeconomic imbalances over time. The contribution of transfers to fiscal imbalances is most evident at the level of the central administration, whose overall position deteriorated significantly between 1991 and 1996. The largest source of growth in central administration expenditure during 1991–96 has been current transfers;4 of this, transfers to the regional and local governments have accounted for a steadily growing percentage. Thus, a great deal still needs to be done to reform the structure of the existing transfer system and to provide lower levels of government with the ability to utilize own resources at the margin, and thus to engender accountability. This involves effective tax administration, as well as appropriately designed transfers, budget, and information systems, together with controls on borrowing.

Colombia has a three-tier structure of government—including the national government, departments, and municipalities—with expenditure responsibilities, as well as own sources of finance for each tier. In Colombia, each level of government is defined as having central, decentralized, and state-owned components. Many of the smaller municipalities do not so far have the capacity to manage own taxes, or to finance current expenditures out of disposable revenues. Many of the larger municipalities have modern systems of tax administration and appear to be in a better position to take on expenditure functions and raise own revenues than most departments. The departments have a limited capacity to take on the expenditure responsibilities in the areas of health and education that are assigned to them, and the national government has created special funds for these purposes, so that the transfers for these expenditures do not even enter departmental budgets. Thus, both departmental and municipal accountability are fairly limited.

Expenditure and Revenue Assignments

Expenditure Assignment

The 1991 Constitution devolved substantial expenditure responsibilities to municipalities and departments, in a manner that is consistent with the benefit principle of expenditure assignment.5 The objective is to achieve outcomes that are responsive to local needs, and thus better targeting as well as efficiency in provision. However, the decentralization process in Colombia is still in its infancy. The departments and most municipalities lack the institutional capacity to effectively perform assigned expenditure functions.6 Decentralization is further constrained by a continuation of the strong tradition of centrally determined norms for expenditures, such as education and health.

Perhaps the greatest constraint facing the effective assumption of expenditure responsibilities at the territorial level is that neither departments nor municipalities have sources of own revenues fully under their control—many of the ceded taxes and transfers are earmarked. Moreover, there is uncertainty concerning the total cost implications of the decentralized responsibilities. This is partly because expenditure policies continue to be determined by the national government, and there is a danger of ending up with unfunded mandates.

The widespread use of earmarking is a result of accountability problems—the national government is not sure that public monies will be used for “appropriate” purposes. In theory, a vertical hierarchy, with a number of special funds and earmarking, should ensure that resources are properly spent according to centrally made decisions. However, decentralized implementation in the absence of proper information flows and budgetary procedures, in a country like Colombia, can severely hamper the task of monitoring outcomes or of assessing policy priorities correctly. As in many countries, arbitrariness and corruption are also possible.

Moreover, centrally determined funding of public functions through earmarked taxes results in budgetary inflexibility and runs the risk of inefficiency and waste. Resources channeled through funded programs may diverge from local needs in the longer run, and it is politically difficult to abolish strings once they have been attached. Moreover, earmarking allows specific interest groups to secure a share of public resources for long periods. And finally, the centrally determined funding of public activities reduces the scope for autonomous public policy at the local level, as it often implies the predominance of national goals. This runs counter to the objectives and spirit of decentralization.

Tax Assignment

The tax assignments of the national government accord largely with basic principles of intergovernmental finance and appropriate tax administration.7 The structure of intergovernmental fiscal relations in the early 1990s is described in detail in Wiesner-Duran (1992). The Constitution of 1991 replaced the municipal share of the value-added tax (VAT) by a share in current revenues (the Participación, described below).

Current System

National Taxes

All major taxes are assigned to the national government, including the VAT, international trade, and personal and corporate income taxes. This assignment corresponds to international practice and also facilitates administrative efficiency—given the functional structure of DIAN (the tax administration and customs department). Altogether, the national government raised about 11 percent of GDP in 1996 (or Col$10,210 billion) in tax revenue from these sources.

The structure of national taxation is shown in Table 1. Income taxes and the VAT contribute roughly equal shares to total national tax revenue (slightly above 40 percent each). Taxes on foreign trade add another 10 percent, and among the remaining taxes, only the tax on gasoline plays an important role, with about 6 percent of total tax collection.

Table 1.Colombia: Selected Tax Revenues at National, Departmental, and Municipal Levels
(In billions of pesos)(In percent of GDP)
National taxes
Income taxes2,3172,8703,3905.15.24.7
Value-added tax2,2412,7493,4334.95.04.7
Trade taxes5527068821.21.31.2
Other taxes3414194840.80.80.7
Departmental taxes
Motor vehicles15955700.350.100.11
Municipal taxes
Industry and commerce1993761,0470.440.661.49
Motor vehicles3751140.080.090.02
Total (all levels)6,7488,2519,95114.8114.6214.15
Source: Ministry of Finance.Note: These data do not include payroll taxes.
Source: Ministry of Finance.Note: These data do not include payroll taxes.

Departmental Taxes8

The national government has ceded all excise taxes on the consumption of liquor, beer, and tobacco to the departments (Table 2). The departments also administer these taxes, with the exception of the beer tax, which continues to be administered by the national government. In addition, the departments administer and collect a tax on motor vehicles (impuesto de timbre), which is similar to a tax at the municipal level with the same base (impuesto de circulatión y tránsito). Furthermore, the departments are assigned a registration tax (impuesto de registro y anotación) on official documents such as inheritances and property transactions (where the tax is based on the intrinsic value of the document), as well as lump-sum transactions such as the certification of diplomas. The registration tax is thus a blend of a stamp tax and a traditional transactions tax and is similar to an inheritance and gift tax, or a property transaction tax.

Table 2.Colombia: Individual Taxes as a Share of Departmental Tax Revenues
Tax revenue10.890.870.560.440.360.580.960.950.95
Vehicle tax0.
Source: Ministry of Finance.

As a proportion of departmental current revenues.

Source: Ministry of Finance.

As a proportion of departmental current revenues.

There are a number of smaller taxes on gambling with a great variety of tax bases, lottery premiums, lottery ticket sales, horse racing, and betting games, each with a complicated set of rules. Similar taxes on gambling exist also at the municipal level. The departments also operate a slaughter tax, again in parallel with municipalities. The revenue picture of the departments is further complicated by the fact that they operate regional monopolies in major sectors. This is true for lotteries, and for the consumption of specific goods, like liquor.

Altogether, the departments raised 1 percent of GDP in taxes in 1995. The structure of departmental taxation is shown in Table 2. Excises provide the major source of revenue to the departments and accounted for more than three-fourths of total tax revenue in 1995.

The liquor tax is particularly important, with nearly 40 percent of total departmental revenue. The other major departmental tax, on the registration of motor vehicles, raised about 10 percent of total departmental tax revenue in 1995. The registration tax, which could be a substantial revenue raiser in view of increasing values for property transactions, contributed only about 2½ percent to total departmental tax revenue in 1995, significantly less than the revenues generated by one of the taxed consumption goods. All other departmental taxes provided about 10 percent of total revenue.

Municipal Taxes9

Municipal taxation in Colombia rests on two main pillars: a property tax (impuesto predial unificado) and a business tax (impuesto de industria y comercio). Each of these taxes contributed from 20 percent to 60 percent of total municipal tax revenue in 1995 (see Table 3). The municipal motor vehicle tax raised about 2 percent of total municipal tax revenue.

Table 3.Colombia: Municipal Taxes(As a share of municipal current revenue)
Tax revenue0.520.510.59
Property tax0.210.200.24
Business tax0.230.210.26
Vehicle tax0.0040.0030.01
Santafé de Bogotá
Tax revenue0.870.850.85
Property tax0.150.140.21
Business tax0.340.340.36
Vehicle tax0.120.150.00
Tax revenue0.340.6110.51
Property tax0.190.370.22
Business tax0.110.210.13
Vehicle tax0.020.040.02
Tax revenue0.210.210.52
Property tax0.110.130.33
Business tax0.060.060.10
Vehicle tax0.006Ü.0060.01
Source: Ministry of Finance.

The large increase is due to a negative number for other taxes, probably a redistribution to other departments.

Source: Ministry of Finance.

The large increase is due to a negative number for other taxes, probably a redistribution to other departments.

Apart from these major taxes, municipal governments in Colombia operate a host of smaller taxes, some of which were introduced on an ad hoc basis. Among these are the slaughter tax, a tax on public performances, a tax on club sales, a tax on the extraction of sand, gravel and stones, gambling taxes, and stamp taxes. Some taxes generate very little tax revenue, complicate the administration, and render the municipal tax system unwieldy.

Among the taxes with a fair revenue-raising potential is the contribución áe valorización. This is a betterment levy (rather than a tax), by which the costs of specific public works projects are shared among direct beneficiaries. This levy is based on the benefit principle and could eventually raise more significant amounts of municipal revenue than in the past, if more widely applied. Another potentially important revenue source could be a surcharge on the price of gasoline. Medellín uses this instrument to finance its metro project, and other municipalities have followed suit recently.

All local taxes raised only about 2 percent of GDP in 1995. Given their much higher expenditure responsibilities (about 6.1 percent of GDP in 1995), municipalities have to rely to a large extent on national support through grants and cofinancing arrangements. Given assured and increasing intergovernmental transfers, as stipulated under the Constitution, municipalities have resorted to borrowing, as discussed below.

Vertical Imbalances

Taxes, such as the VAT, and the corporate and personal income tax, are assigned and administered at the national level as in many countries. This leads to the problem of vertical imbalances, as illustrated in Table 4. For example, in 1994 municipal government tax revenues were only 32 percent of municipal expenditures, while national government tax revenue was 87 percent of national expenditures.

Table 4.Colombia: Vertical Fiscal Imbalance Between Levels of Government, 1994
In Billions of PesosIn Percent of Own Level ExpendituresIn Percent of Total Expenditures
National level
Own revenue9,549100.980.1
Nontax revenue9319.86.5
Departmental level
Own revenue1,00163.78.4
Nontax revenue28718.32.0
Municipal level
Own revenue1,36741.511.4
Nontax revenue3699.72.6
Own te venue11,91883.2
Nontax revenue1,58711.1
Source: Ministry of Finance, National Planning Department.Note: Expenditures include operational surplus or deficits of public enterprises at each level. Expenditures do not include transfers to other levels of government. Net outlays of public enterprises are consolidated with the expenditures at each level of government.
Source: Ministry of Finance, National Planning Department.Note: Expenditures include operational surplus or deficits of public enterprises at each level. Expenditures do not include transfers to other levels of government. Net outlays of public enterprises are consolidated with the expenditures at each level of government.

Despite the vertical imbalance in its favor, the national government has been increasingly constrained by the constitutional stipulation on increasing transfers to the territorial governments. However, since decentralization is not yet fully functional, the national government continues to bear expenditure responsibilities that have not been effectively devolved to territorial governments (for example, teachers’ salaries). On the other hand, Table 4 may understate the extent of the potential vertical imbalance, given the degree of earmarking that exists in Colombia—that is, mandated expenditures that have to be met out of the own resources of departments. Part of the difficulty is that the national government retains control over overall wages and salaries, which together with other contingent liabilities and the limited own resources (own revenues plus untied transfers) of the local governments, make it difficult for the local authorities to assume the expenditure responsibilities fully.

Problems Associated with Current Revenue Assignments

The tax assignments of the territorial governments are designed to strengthen their autonomy based on own resources. However, own revenue collections of departments and municipalities combined represented a little over 3 percent of GDP in 1995. The social and economic conditions among the different regions of Colombia vary widely, as reflected in GDP per capita among departments. The standard deviation of per capita income among departments is 42 percent of the mean value, which is high compared with industrial countries. The dispersion of per capita income affects regional taxable capacity, as well as the differentials in basic needs for public services of departments and municipalities.

Recent legislation assigns royalties on natural resources (such as petroleum) mainly to producing municipalities and departments. Despite the redistributing role of the Fondo Nacional de Regalías, horizontal fiscal imbalances are likely to have been exacerbated by this legislation, with associated efficiency costs in the longer run.

Horizontal Tax Competition

Colombia faces a number of problems relating to horizontal tax competition, at both the departmental and the municipal level. The departmental system of specific consumer taxes generates domestic contraband of taxable goods (such as liquor and tobacco). The local surcharge on gasoline consumption also encourages tax competition. A redesign of these taxes would be necessary to reduce illegal trafficking and easy avoidance, and to facilitate administration and increase revenue capacities.

Intergovernmental Transfer System

To correct the vertical imbalance, a system of national government transfers to departments and municipalities has been used since the late 1960s.10

Tax Sharing

In the past, Colombia shared the VAT with municipalities, but now practices tax sharing on a more comprehensive base. The sharing of a comprehensive revenue pool avoids several disadvantages that result from sharing separate revenue bases (as described above). However, the shared revenue base is not as comprehensive as might appear at first sight. For instance, for the first year in which they are in force, the national government does not include in the pool surtaxes on the income tax, the resources resulting from broadening the base of shared taxes, and increases in the consumption tax.11 To introduce greater flexibility to resource allocation and use, tax sharing should be untied. The a priori earmarking of tax revenue causes complications for a revenue-sharing scheme.

There are four main mechanisms for revenue sharing between the national, regional (departamentos), and local governments (municipios): (1) the situado fiscal, which provides automatic transfers to the regional governments earmarked for current expenditure in health and education; (2) the participación municipal, which earmarks transfers to the local governments for current and investment expenditure on basic services including education, health, and water supply; (3) the system of cofinancing funds; and (4) the National Royalties’ Fund, Under the first two mechanisms, all central government tax and nontax revenue is subject to revenue sharing. As required by the 1991 Constitution, which mandated steady increases in transfers to the end of the decade, revenue shared with the regions under these two systems has increased steadily in the last four years as a share of central government current revenue (Table 5). In 1991, approximately 12 percent of the central government’s current revenue was transferred to the local governments;12 this rose to 18 percent in 1997, and is scheduled to reach 22 percent by 2001. In 1996, central government revenues transferred to both the regional and local governments accounted for 41.5 percent of total central government current revenue.

Table 5.Colombia: Transfers of Central Administration Current Revenue(In percent of total current revenue)
To municipal governments
To regional governments
Source: Ministry of Finance.
Source: Ministry of Finance.

Since 1993, the size of the situado fiscal has been set as a minimum share of the central government’s total current revenues, and has varied from 23 percent in 1994 to 24.5 percent in 1996. The 1991 Constitution stipulates that neither the situada nor the participaciones (regional and local transfers) can diminish as a share of central government current revenue in relation to the previous year’s share. Moreover, in the event of a shortfall in the national government’s current revenue, transfers cannot be automatically reduced in the same proportion. This inflexibility in budget rules significantly constrains the fiscal management of national revenues.

Transfers to Departments: Situado Fiscal

The 1968 Constitution introduced the situado fiscal, a system of transfers that has continued to be the central pillar of departmental finances. The basic features of the situado remain as the fundamental precepts of the current transfer system. The distribution formula was as follows: 30 percent of the total revenue was distributed in equal shares among departments, territories, and the national capital of Bogotá and the remaining 70 percent was distributed according to the population.

The transfer was earmarked for health and education. At the same time, a complete reorganization of the provision of these services was made through the creation of the Regional Education and Health Funds (Fondos Educativos Regionales and Fondos Seccionales de Salud), mainly dependent on the national government. These agencies received direct payments from the situado fiscal, without a line item in the budgets of the departments. In other words, the role of departments in their two most important areas of responsibility was reduced to supplementing funds, if any, from their own sources to these agencies.

Law 60 of 1993 provides additional legislation on transfers, pursuant to the 1991 Constitution: the total amount of the situado will be determined as a minimum share of the nation’s total current revenues, ranging from 23 percent in 1994 to 24.5 percent in 1996. A new law is thus needed for 1997. Law 60 defines the distribution formula for the situado as follows: 15 percent of the total revenue is to be distributed in equal shares among departments, territories, and the national capital of Bogotá and the remaining 85 percent is to be distributed according to the population. This distribution clearly favors the smaller departments, and may have some net distributional impact. For example, in 1995 the smallest departments received the largest transfers in per capita terms for health and education. In the same year, the largest share of transfers earmarked for education went to finance primary and secondary education, as determined by Law 60.13 The distribution criteria set by Law 60 have varied from year to year to ensure a smooth transition from the previous system to the new one, which was originally scheduled to be in place by 1997. However, during the first half of 1997, no legislative initiatives had been taken to determine the amount of the increase (or decrease) in the situado or to redefine distribution criteria.

As mentioned above, departments may use the situado for education and health only. According to Law 60, departments are required to devote at least 60 percent of the situado to education and 20 percent to health. Departments may at their discretion allocate the remaining 20 percent according to their priorities, but only within these two sectors. In effect, the local discretion is limited, and the situado is essentially a special purpose transfer. From the departments’ perspective, there is a reluctance to assume directly their expenditure responsibilities since they have no clear perception of their future spending obligations.14 The problem is particularly acute in education.

Transfers to Municipalities: Participación Municipal

The second major transfer scheme is the participación municipal, a special purpose transfer to the local governments established by the 1991 Constitution to finance recurrent and investment expenditure for basic services. Between 1968 and 1991, these transfers were general purpose transfers to the municipalities. An important restriction in the Constitution is that the transfers are to be determined on the basis of total current (tax and nontax) revenues of the national government,15 and that the share of these revenues going to local governments is to increase from 14 percent in 1993 to 22 percent in 2002. The Constitution also stipulates that the transfers be distributed 60 percent according to the number of poor people and people with unsatisfied basic needs; the remaining 40 percent according to population, “fiscal and administrative efficiency,” and progress made by municipalities in improving the quality of life in their territory. Again, Law 60 of 1993 specifies the formula and determines the share of current revenues to be distributed each year. It also sets the criteria for the use of funds.

During a transition period for the distribution of participación transfers, every municipality received annually a basic transfer, equal, in constant prices, to the transfer received in 1992. The remaining amount of the transfer was distributed as follows: 40 percent in direct proportion with the number of persons with unsatisfied basic needs; 20 percent according to a poverty index; 22 percent according to population; 6 percent according to a fiscal efficiency index, measured by the increase over two years of per capita municipal own revenues weighed by a relative unsatisfied basic need index; and 6 percent according to an administrative efficiency index, measured by the ratio between the rate of increase over two years of current expenditures (other than on education, health, and infrastructure) and the number of inhabitants with water and sewer services. A special provision was made for small municipalities (less than 50,000 inhabitants), by allocating to them a share of 5 percent of the total transfer, net of its basic component.

Participación funds are earmarked for the following uses: 30 percent for education; 25 percent for health; 20 percent for water provision, in case the share of population serviced is less than 70 percent; and 5 percent for sports, recreation, and culture.

A transition period was introduced by Law 60, according to which municipalities were given time to adjust the structure of their expenditures to this norm.16 Since the 1995 ruling by the constitutional court declaring this norm unconstitutional, a number of municipalities have been unable to finance their operating and debt-servicing expenditure out of own tax resources.

The Cofinancing Funds

To complement the system of specific purpose grants described in the sections above, a system of cofinancing funds was gradually introduced over the years.17 Cofinancing resources were expanded until 1996, and in 1995 represented almost 40 percent of transfers to municipalities. As in the case of other countries, specific purpose grants allow the donor, here the national government, to pursue specific priorities in the various fields. These transfers are made mostly on a matching basis. Providing technical assistance to beneficiary governments has also been an important function of the cofinancing arrangements.

Until recently, the main cofinancing funds have included: the Fondo de Cofinanciación para la Infrastructura Vial, for intermunicipal roads; the Fondo de Cofinanciación para la Infrastructura Urbana, mainly for municipal transport; the Fondo de Inversión Social, for social investment projects (such as education); and the Fondo del Desarrollo Rural Integrado, for infrastructure projects in rural areas. The principles for distribution of the cofinancing resources are similar to those of other transfers, including the number of persons with unsatisfied basic needs in each department and the share of other transfers going to the municipalities in each department. A needs indicator is combined with a fiscal capacity indicator. There are different matching shares, for which recipients may use either their current revenues or borrowed resources, ranging from 10 percent for poor municipalities to 40 percent for rich municipalities and to 50 percent for big cities.

While beneficial to the local governments, the cofinancing system has had important flaws. These are related to the lack of coordination, and the ad hoc objectives assigned to the funds. Annual programs are independently formulated by each agency, with indeterminate criteria for project selection and cost sharing. There are no consistent directives for guiding the agencies in their operation. Moreover, the grants are distributed among various regions more in order to correct perceived weaknesses of the transfer system than to implement national priorities in given fields of activity. In some municipalities and departments, monies allocated for cofinancing in some fields, such as preschool education, are not used. This is partly an information problem, particularly in the smaller municipalities. In other areas, projects are formulated solely to obtain access to financing.

In contrast to the situado and the participaciones, the 1991 Constitution does not stipulate that the cofinancing funds must increase at a certain rate in relation to central government current revenue. Therefore, the national government has much greater flexibility in determining the level and allocation of these funds to the regional and local governments. As a reflection of this flexibility and as a result of recent measures to reduce 1997 budgetary outlays, the cofinancing system has recently undergone significant changes. In early 1997, a series of measures were announced affecting the cofinancing system. These measures included: (1) a change in the basis of allocation of funds from grant funds to loanable funds on concessionary terms; (2) a significant reduction in the budget allocation for cofinancing; and (3) a reduction in the number of cofinancing funds to one.

The National Royalties’ Fund

The National Royalties’ Fund also represents an important source of revenue for the local governments, particularly in the wake of the large oil discoveries in the early 1990s.18 Royalties distributed in 1995 were nearly equivalent to the cofinancing funds. This Fund, established by the 1991 Constitution and further regulated through legislation passed in 1994, was set up to distribute natural resource royalties primarily to the producing regions for purposes of investment.19 The distribution of royalties among the various regions is on a presumptive equal per capita share, but the skewness of the actual distribution is striking. Despite the right of producing departments for a claim on royalties generated in their territory, small departments with a tiny population, like Arauca, get a large amount of royalties, giving ground to charges of misuse of resources.

Evaluation of the Present System

The overall distribution of transfers is shown in Table 6. Leaving aside royalties, transfers seem advantageous to departments with low per capita GDP (these are marked with an asterisk in Table 6). However, the overall redistributive impact is not clear. Rich departments include in their jurisdictions the biggest cities, with a high concentration of poor migrants and of unemployment. The distribution of royalties has a major impact on the distribution of the entire system of transfers to territorial governments. Small oil producing departments benefit from higher per capita sums than big departments, further clouding the overall effect of transfers and royalties.

Table 6.Colombia: Transfers to Departments and Municipalities(In billions of pesos unless otherwise indicated)
Situado FiscalTransfers to MunicipalitiesRoyaltiesTotalPer Capita (In pesos)
Barranquilla (Distr.)22.935.716.727.83.60.0643.363.743,83556,017
Santafé Je Bogotá (Distr.)163.2258.274.7121.717.5255.5380.054,18169,296
Cartagena (Distr. of)17.448.610.518.92.41.530.369.046,61286,944
Casan are10.818.06.514.527.5101.544.8133.9267,588633,818
La Guajira18.933.910.917.810.021.139.972.8119,925168,199
Norte de Santander53.288.328.950.23.63.485.7141.987,855122,114
San Andrés5.810.,426202,699
Santa Marta9.924.,05698,835
Valle del Cauca115.5189.072.3118.912.70.42200.6308.362,064482,543

Links Between Revenue and Expenditures

Decentralization in Colombia has clearly enhanced the revenue side of the budget of territorial governments. The overall sense of constitutional provisions in this field is that the nation considers allocating to territorial governments a large and rapidly growing share of its own revenues as an essential step for putting the decentralization process on the right path. Devolution of expenditure responsibilities has been mostly left to ordinary legislation. This weak correspondence between revenue and expenditures appears to be a major problem for the overall equilibrium of the public sector, especially if account is taken of the fact that territorial governments are encouraged to use their transfers as a leverage to increase their reliance on debt financing.

The share of the nation’s current revenues allocated to territorial governments is presently close to 50 percent and will grow if the trend is to proceed along the guidelines provided by the Constitution. However, these percentages are typical of fully fledged federal systems, where responsibilities of subnational governments are much wider. The earmarking of these transfers appears to hinder effective decentralization. Thus, the central government continues to bear much of the expenditure responsibilities, although resources have been transferred to territorial levels of government.

Constraints on the Use of Transfers

While territorial governments now have more resources to spend than under the previous system, discretion in the use of funds has been even more curtailed by fixing distinct percentages to be spent for every function. The Colombian approach to decentralization seems to still favor the model of functional administration, typical of highly centralized systems, in which the center sets the priorities and decentralized institutions are mainly asked to implement them.

There are clear reasons for this approach, but the problems are also evident. On the one hand, newly elected mayors, governors, and local councils frequently lack administrative experience, political and bureaucratic misuse of resources is still widespread, and control by voters is still in its infancy, particularly in the rural areas. On the other hand, excessive constraints may be counterproductive. They tend to reduce responsibility at the local level, when, as is usually the case, local priorities diverge to a certain extent from national priorities. In fact, one of the main advantages of decentralization is the geographical differentiation of policies, according to locally perceived priorities and voters’ preferences in the sphere of autonomy delegated to territorial governments.

When rigid constraints impede satisfaction of first-level priorities, local authorities will turn to second- or third-level priorities. School buildings will be constructed when there is no additional need, whereas the extension of aqueducts is badly needed, since tap water covers only a tiny percentage of the population. Rigid, detailed rules are not followed, weakening cooperation and coordination between national and local authorities. The former are then tempted to issue new, even more rigid, constraints.

In all countries, there are national priorities within the spheres left to local authorities. These priorities may be better pursued by using specific purpose grants, with simple criteria, rather than complicated criteria (more suited for general grants), which are difficult to monitor.

The Emergence of New Territorial Rents

Although the 1991 Constitution set up a framework for a more equitable distribution of royalties, legislation seems to have ceded excessive ground to pressures coming from producing areas. No basic change has been introduced over the previous allocation system, since producing departments and municipalities still retain their shares of the royalties within realistic levels of production, and new rents have been created for ports and other areas marginally affected by the production of natural resources. Whereas there is certainly an argument for compensating local jurisdictions, where production and transportation of natural, polluting resources takes place, the distribution rules for royalties go beyond this necessity, depriving the nation of an important source of revenues for the implementation of its priorities. Moreover, royalties exert a distorting effect on the overall design of transfers, creating areas of privilege and inefficiency.

Problems Related to the Formulas

A number of countries, particularly in Latin America, rely on special purpose transfer mechanisms similar to those examined in the Colombian context. Some of the shortcomings of such transfers are discussed below.

Incentives for Fragmentation

The 15 percent “equal share” provision of the situado fiscal provides an incentive to create new jurisdictions. A similar effect is induced by the special provision for small municipalities, jointly with the rapid increase of the total amounts transferred to them. International experience shows that the positive effects of decentralization do not derive from the number of local jurisdictions, but from their ability to provide public services to their citizens, which implies a minimum size for efficient operation of local governments.

Inadequate Information

A problem frequently found by countries using formulas, as in Colombia, for funding territorial (especially municipal) governments is the lack of up-to-date information. Colombia seems to present many problems in this respect. Population and other data used for the formulas are available only at census years. The high mobility of the Colombian population makes census data projections problematic. The last census took place in 1993, and total population figures were published in 1997. From 1985 to 1997, the formulas utilized (controversial) projections for population, based on data from the previous two censuses for the most important indicator, that is, the number of poor and of persons with unsatisfied basic needs. Thus, the 1985 census has been used to estimate this number,20 and the rates of change between the 1973 and 1985 censuses were used to project changes to the current period. This favors static areas, or areas with a declining population, and penalizes those areas, including the big cities, experiencing massive inflows of new migrants.

The Use of the Same Indicators for Different Transfers

The same indicators, or slightly different versions thereof, are used for the distribution of transfers.21 This contradicts one of the basic rules for economic policy, according to which the number of instruments should be correlated to the number of objectives. In other words, if every category of transfers uses the same criteria for allocating funds, there is no need to use different transfers—one grossed up would be enough. If the nation really desires to pursue distinct specific objectives with its transfer system, it should have recourse to distinct indicators for each category of transfers, properly targeted to its aims and priorities.

Local Tax Administration

The problems discussed above with the structure of the existing revenue-sharing system must be viewed in the context of the local governments’ capacity to raise and manage their own revenue. The structure of the departmental and municipal tax systems is discussed in greater detail in Appendices I and II, respectively. This section focuses on the status of tax administration at the local level.

In Colombia the capacity of departmental and municipal tax administrations varies widely. A few tax administrations, usually in large municipalities, are in the process of establishing modern procedures. However, many tax administrations at the departmental level, and in the smaller municipalities, lack the capacity to perform basic tax administration functions in an effective manner. Almost half (44 percent) of the 1,048 municipalities in 1993 had a population size of less than 10,000. For these municipalities, there are certain diseconomies of scale associated with developing a modern tax administration. Summaries of tax administration procedures in 1994, for selected departments and municipalities, are presented in Tables 7 and 8, respectively.

Table 7.Colombia: Tax Administration Functions for Selected Departments
Number of tax staff19610292
Organizational structure
By collection and enforcement functionsXXX
Annual plan
Taxpayer services (information, assistance, and education)
Payment at banksXX
Payment at tax officeXX
Specific tax sanctionsXXO
Payment incentives1OXO
Source: Discussions with tax officials.Note: X = exists; — = does not exist; and O = unknown.

Tax officials receive a share of the cash value of goods confiscated.

Source: Discussions with tax officials.Note: X = exists; — = does not exist; and O = unknown.

Tax officials receive a share of the cash value of goods confiscated.

Table 8.Colombia: Tax Administration Functions for Selected Municipalíties
Number of tax staff24030037384
Number of taxpayers
Property tax1,400,000300,00051,00030,00010,400
Industry and commerce tax56,00050,00010,0001,110120
Organizational structure
By functionX
By type of taxXXX1
Annual planXXX
Taxpayer services (information, assistance, and education)
Payment at banksXX3XX
Detection of stop filersXXX
Number of auditors51720500
Skill levelmediummediumlown.a.n.a.
Self assessment
Property taxX
Industry and commerce taxXXXXX
Source: Discussions with tax officials.Note: X = exists; — = does not exist; O = unknown; and n.a. = not applicable.

In Sasaima, the tax office is part of the Treasury Department.

Accounts of taxpayers for both the property tax and the industry and commerce tax have been computerized by contracting out to firms that specialize in this technology.

Payment is also allowed at the tax office, by cash or credit card.

The municipal taxpayer identification number for business is the same as the national taxpayer identification number, which facilitates cross-checking with DIAN.

In Bogotá another 40 auditors, and in Medellín another 30, were available on a contractual basis.

Source: Discussions with tax officials.Note: X = exists; — = does not exist; O = unknown; and n.a. = not applicable.

In Sasaima, the tax office is part of the Treasury Department.

Accounts of taxpayers for both the property tax and the industry and commerce tax have been computerized by contracting out to firms that specialize in this technology.

Payment is also allowed at the tax office, by cash or credit card.

The municipal taxpayer identification number for business is the same as the national taxpayer identification number, which facilitates cross-checking with DIAN.

In Bogotá another 40 auditors, and in Medellín another 30, were available on a contractual basis.

Taxpayer Services

The large municipalities address the tax administration function of providing such taxpayer services as information, assistance, and education in varying degrees. Officials in these municipalities are well aware of the importance of taxpayer services. In small municipalities, however, the capacity does not exist to provide a full range of basic services.

In departments too, only a few of the basic services are provided. Taxpayers are forced to wait in line to obtain necessary information. In some cases, especially with respect to the excise taxes, taxpayers must follow several time-consuming steps before payment is completed. Many of the officials do not appear to be concerned about taxpayer inconvenience.

An important component in the strengthening of tax administration is awareness of the taxpayers’ perspective. Important steps include the development of a more comprehensive taxpayer services function and the streamlining of procedures to reduce the compliance burden on taxpayers. In addition, various tax forms should be simplified as much as possible.


The collection process, in general, is relatively strong in the local tax administration in Colombia. Although some medium-sized municipalities have not computerized collections, most small municipalities have received computers and special software (SIAM) from the national government. Many tax administrations have developed units that monitor the payments of large taxpayers, since a sizable percentage of the revenue comes from a small number of large taxpayers. For those tax administrations that cannot quickly detect stop-filers, a first step would be to identify the large taxpayers (say the top 10 percent) and begin to monitor (even on a manual basis) their payments.


Small municipalities and departments lack an audit function. Departments focus on deterring smuggling as their primary enforcement tool. They do not focus on the taxpayer. Small municipalities have little capacity to question the information provided by the taxpayer on the industry and commerce tax return. As the decentralization process unfolds, it will be in the best interest of each level to establish an audit function. A first step may be to follow the action taken in large municipalities, which is to hire contractual auditors for specified periods of time.22 Regular staff could observe these contractual auditors performing desk audits, and accompany them on field audits to learn how to conduct an audit. Additional training is another step. Observance of audits should be accompanied with classroom training in accounting and audit techniques.

Audits alone will not create a climate of voluntary payment of taxes. Taxpayers must know that there is a risk of detection if they fail to register, fail to file a return, or fail to pay the correct amount on time.

Cross-checking of information with other sources through use of the computer is an important enforcement tool. Some municipalities have cross-checked information with the national tax administration (DIAN). But this effort has largely been on a case-by-case basis, rather than a systematic initiative that would encompass all taxpayers. The municipalities visited are interested in systematic cross-checking and plan to explore with DIAN ways to achieve this objective.

Penalties and Appeals

A speedy and well-designed appeals process protects taxpayers’ rights, while complementing the penalty structure. Small municipalities do not seem to be able to apply penalties. They try to provide an incentive by discounting the tax bill if paid by a certain date. They lack a simple system of penalties and appeals, along with the experience to design and implement such a system. Without such a system and the resolve to apply it, the tax administration in small municipalities will be unable to raise adequate own source revenue.

Appeals can be resolved within two years at the national level, but appeals at the departmental and municipal levels go to the civil courts, and may take up to ten years to resolve.

Other Issues

Development of an Annual Plan

The larger municipalities have a planning process, and Bogotá has a very sophisticated one. However, the smaller municipalities lack a clear sense of priorities and goals. It is important for staff in these tax administrations to develop a better perspective of the overall goals of tax administration, and how their specific tasks help to attain the goals. Development of an annual plan can be an initial step in a process to modernize tax administration at the municipal level.

Dissemination of Information

Given that tax administrations at each level deal with a common set of concerns regarding taxes at that level, it would be important to develop a network to share innovative ideas. For example, large municipalities could meet with medium-sized municipalities to explain how they have implemented modern procedures. The medium-sized municipalities could then, in turn, meet with the small municipalities to discuss administration advances they have made. In order to realize an organized dissemination of information, which would include techniques to develop an annual plan, a coordinating body is needed which could profit from technical support from the national tax administration.

Adjustment for Inflation

One final tax administration issue to consider is the problem of collection lags. Currently, municipalities, except for Bogotá, collect the industry and commerce tax based on annual sales of the previous year. With inflation, the real value of the taxes collected is reduced. One way to address the problem of collection lags would be to increase the sales of the previous year by expected inflation for this year. Another way would be to apply periodically an inflation adjustment mechanism to tax liabilities.

Territorial Debt and Expenditure Management

As observed above, earmarked own revenues, with nationally determined use of transfers, limit the effective decentralization of expenditures envisaged in the 1991 Constitution, and encourage the use of credit. Banks have an incentive to lend on the strength of constitutionally increasing transfers from the nation. While there is concern that growing territorial borrowing could pose a threat to the country’s macroeconomic stability, the data available from various sources are partial and inconsistent.

Weaknesses in expenditure reporting and control mechanisms at the territorial levels severely hamper prospects for “accountability” in subnational spending. Progress in this area is of high priority if the full benefits of decentralization are to be realized.

Territorial Debt

Territorial borrowing by the administrations at each level increased rapidly in nominal terms from 1993 to 1996 at an average annual rate of 67 percent for the departments and at a rate of around 55 percent for municipalities. When public sector enterprises are included, the overall growth rate of territorial debt was lower at 16 percent for 1994. During the same period, the stock of debt nearly doubled, from around 1 percent to nearly 2 percent of GDP in 1996.

It is curious that departments, in particular, appear to have maintained positive net asset balances with the banking system at a time when borrowing has increased sharply. This paradoxical situation is due to the high degree of earmarking of transfers from the national government and of own revenues, together with the legal requirements to maintain separate accounts for each source. Given weaknesses in monitoring, commercial credit acts as a more “fungible” resource for the territorial governments.

Budgets for three municipalities and two departments for 1995 show that in each case there have been increasing deficits, with inelastic revenue sources and additional expenditure responsibilities associated with the increasing decentralization envisaged under the 1991 Constitution, as well as contingent liabilities under the recent pension reform plan.

The government’s macroeconomic plan for 1994–98 is predicated on a central government deficit counterbalanced by surpluses at the territorial and public enterprise levels. However, as a whole the territorial governments have been in a deficit position from 1994 to 1996, and the prospects for improvement in the near future are not encouraging, given present trends.

Policy Initiatives

The Ministry of Finance has taken several measures that should effectively curtail the rate of increase in territorial borrowing. First, a decree was issued in mid-1995 increasing the amount of revenues that might be used as collateral, from 130 percent to 150 percent of the guaranteed loans. Moreover, credit institutions have to increase the ratio of net worth to loans to territorial entities. While increasing the cost of credit, these measures should reduce both the supply of credit, as well as the demand (although one-term mayors and governors do not appear to be particularly interest-sensitive).

Second, a new law on local government borrowing (Ley de Endeudamiento Territorial) was approved in early 1997. While falling short of requiring an independent evaluation of the borrowers’ credit risk directly, the law establishes a system of warning signs based on stock and flow indicators of the regional and local governments’ repayment capacity, and specifies the jurisdiction that must approve the requests. However, under the law’s new rules some local governments have actually increased their borrowing capacity. The two indicators are (1) interest payments as a percentage of (current) savings; and (2) the stock of debt as a percentage of current revenue. In addition, local governments cannot borrow to cover current expenditure. The agencies designated to supervise the new borrowing rules are the banking superintendency, the Ministry of Finance, and the commercial banks.

At the same time, the Ministry of Finance has begun to implement a program aimed at improving local government fiscal and debt management. The program requires local governments with weak finances to undergo an adjustment process that might include freezing salaries and undertaking other austerity measures. However, the main lacuna in the proposed measures is that the national government does not have the means to implement the controls—given sporadic flows of information and nonexisting external audit mechanisms. Thus, reforms in the management of debt and public expenditures are extremely urgent.

Management of Debt and Public Expenditures

Given that there are over a thousand municipalities, considerable variation in expenditure management practices is possible. However, an examination of the current budgetary situations and practices in some of the largest and more modern municipalities, as well as small and more remote ones, and two departments suggests that even the more advanced administrations suffer from serious shortcomings.

Cartagena and Manizales are among the larger municipalities, and do not lack qualified staff or management capabilities. Manizales, in particular, and Bogotá have made impressive improvements in their tax administrations in recent years. However, difficulties in the public expenditure management aspects are apparent in all municipalities, and these are magnified in the smaller municipalities. Expenditure reporting and control weaknesses in the municipalities severely constrain the effectiveness of the decentralization process.

The territorial authorities face a number of vastly differing reporting formats that are required by a number of national government agencies. In many cases, these reporting requirements lead to a paralysis of the capabilities of the municipalities, with a resulting deterioration in the quality of the information generated. The differing formats also detract from the assessment of the budgetary situation of the territorial governments—needed for an effective management of territorial public finances.

Some of the “standard” formats preclude a meaningful assessment of the overall budgetary situation of individual territorial governments. For instance, there are problems in the classification and formats used for municipal budgets, with an unfortunate confusion between revenues and financing items. Thus, municipalities systematically lack a clear perception of their overall deficit outcomes or prospect and are unable to gauge the extent of their difficulties.

An even greater difficulty is the absence of reliable information on budgetary outcomes. In Cartagena, even cash spending data are seen as unreliable, an issue complicated by the legal requirement to maintain unrelated separate funds and earmarked revenues. The Treasury does not keep a register of spending, and consequently does not have an accurate picture of the overall level of cash spending, or of its destination.

The problem of expenditure monitoring is exacerbated at all municipal levels by the absence of “impartial” and technically qualified controllers. Existing controllers are appointed by local legislatures and are subject to bias. The smaller municipalities do not have adequate own tax revenues to qualify for a controller, and reliance on the departmental controller is tantamount to the absence of any form of control or accountability.

Thus the absence of verifiable expenditure outcomes is likely to have repercussions on possibilities for restraining territorial indebtedness (for example, through measures such as “current savings” that rely on expenditure outcomes as well as revenues raised). All the municipal and departmental data suggest inconsistencies between the financing available and the implied primary and overall deficits.

Departments tend to have weaker management capabilities than the larger municipalities or districts (for example, Bogotá). Thus, the problems seen in the latter level of government are magnified at the departmental level. As in the municipalities, a particular lacuna exists in the inadequate generation of information on expenditure outcomes. And with the inability of the territorial governments to monitor actual spending, the systems of control are nonfunctional. In the departments, as in the case of municipalities, attempts to tie resources and transfers to particular ends because of weak local controls may prove ineffective precisely because relatively stronger controls and reporting mechanisms are required for special purpose grants and tied revenues. Paradoxically, the existence of such “tying” of resources, through special bank accounts, may actually impede the introduction of effective local controls or effective fiscal management, since the overall expenditure situation is not transparent.

Prospects for Territorial Governments

The dynamics of the decentralization process, with weak expenditure management, inadequate incentives to raise own revenues, and easy access to debt finance do not bode well for macroeconomic stability in Colombia.

Banks are keen to lend, given that many territorial governments have earmarked resources that they are unable to utilize together with assured increases in transfers and royalties. And loans to territorial governments are perceived as sovereign debt.

On their part, the territorial governments are keen to borrow, since this provides a potential source of “untied finance.” Also, with the three-year electoral cycle, with no reelection possible for mayors and governors, there is relatively little concern about the consequences of borrowing—this will be someone else’s problem. Indeed, the assessment of budgetary profiles for selected territorial governments bears this out. In Manizales, the stock of debt increased from 69 percent of current revenues in 1994 to 79 percent in 1995, with interest payments increasing by nearly 80 percent in nominal terms in 1995 over 1994 levels. The municipality in 1995 was close to breaching the 30 percent debt service to current revenues criterion. Cartagena, on the other hand, had a stock of debt of 150 percent of current revenues in 1993, and this had risen to 180 percent in 1994. The 1995 budgets indicated that the financing gap may be more than twice as large as current revenues, but because there had been an aggressive refinancing campaign (as well as an indeterminate buildup of arrears), the municipality was comfortably placed vis-à-vis the debt service to current revenues criterion.

The national government was thus correct in seeking to establish a more stringent criterion of unsustainable debt increase at the territorial level, based as a proportion of disposable savings. This indicator would indeed bring attention to bear on the problem territories before the financing difficulties become unsustainable. However, to make the limit functional, improved information systems are needed, vis-à-vis both the banking system and larger municipalities, as well as the monitoring of budgetary trends at the territorial level. Limits to territorial borrowing, together with less earmarking of transfers and enhanced local tax administration, are also crucial in ensuring that greater accountability accompanies the decentralization process.

Appendix I. The Departmental Tax System

Current Taxes

Excise Taxes

The departmental tax system rests mainly on specific taxes on local consumption. Revenue from excises is also earmarked, which reduces the scope for autonomous policy at the departmental level.23 Excise taxation at the departmental level requires the definition of domestic “exports” and “imports” of taxable goods, such as liquor and tobacco, which in turn necessitates internal “fiscal borders” for appropriate monitoring. Moreover, intraregional purchases of taxable goods can be effected only on the basis of mutual agreements among departments. This generates legal and administrative complexities. Although producing factories keep a record of their deliveries to their customers by department, this does not guarantee that the “imported” goods will be taxed by the consuming department. It creates scope for “internal contraband,” which is at odds with the notion of a common Colombian market.

The problem is exacerbated by foreign competition for excisable goods. The importer is required to pay the excises to the department in which the imported goods are sold,24 but there is large scope for tax evasion. Thus, foreign contraband competes with local production of excisable goods. Furthermore, excisable goods are “exported” to other departments “duty free” and this may de facto constitute a loophole through which the goods are eventually “reimported” to the department without the payment of excises.

In order to reduce the incentive for smuggling, the national government in 1994 lowered the tax rate on tobacco, simultaneously creating a Fondo Tabacalero de Compensación Tributaria in order to indemnify producing departments for the loss of revenue. This experience illustrates the fundamental deficiencies of a system of decentralized excise taxation that remains subject to national policy decisions.

The system was further complicated by the fact that the departments were given regional monopolies on the introduction and the sale of excisable goods, for instance for the distribution of liquor (Decree 1222/86, Article 123). This allows the departments to benefit from a monopoly rent as well as from the excise. Recent developments illustrate that regional monopolies and regional taxation of consumer goods cannot exist without the support of national government policy. In order to protect monopoly rents, the national government had to impose customs tariffs on imports. Moreover, the revaluation of the peso, relative to the U.S. dollar, rendered imports more competitive in the Colombian market, putting pressure on monopoly rents and affecting the level of excises.

Other issues of excise taxation in Colombia relate to the tax base and to the rate structure. The tax base for liquor, for instance, refers to the average value of a 750 cc. bottle of spirits sold in Colombia. This value is uniformly fixed by DIAN, the national tax administration and customs department and is updated regularly according to market developments. The tax base is thus a quantity independent of the commercial value of a specific product. A bottle of imported high-quality liquor bears the same tax as a cheap bottle of domestic liquor.

Furthermore, a relatively high tax rate of 35 percent applies to liquors, while aperitivos (aperitifs) carry a much lower rate of 10 percent. The delineation of these product categories by alcohol content (above and below 28 degrees vol.) is arbitrary. The maximum alcohol content of aperitivos is also high, and this induces the production of liquor up to the critical mark, which may not be in the public interest.

The Motor Vehicle Tax

The Colombian departments operate a motor vehicle tax in parallel with the municipal impuesto de circulación y tránsito, using essentially the same tax base. This base is the commercial value of the motor vehicle as established annually by the Instituto Nacional del Transporte (INTRA). The rate structure is progressive and certain exemptions apply (vehicles of the public service and public law enforcement, school buses, agricultural machinery, and so on). Again, a large part of revenue from this tax is earmarked.25 The revenue potential of the motor vehicle tax is substantial. The complexities resulting from the assignment to two tiers of government, with two tax administrations managing the same tax base could, however, be resolved by consolidating the tax at the department level, through a piggybacking or a revenue-sharing mechanism.

Some countries use the motor vehicle tax to monitor and discourage pollution and noise, and in Colombia, the tax is lower on older cars. However, the commercial value of a vehicle does not reflect relevant environmental criteria. Carbon emissions by older cars, with a lower commercial value, may be higher than that on new cars, and a reduced tax on older cars is not warranted on ecological grounds.

The Registration Tax

The registration tax (impuesto de registro y anotación) is levied on the issuance of official documents and the certification of legal acts. The law defines a great number of such acts, some of which are taxed on a lump-sum basis (like a fee), while others are taxed on their intrinsic value. The latter renders the registration tax similar to taxes on gifts and inheritances, for instance, or to taxes on the transaction of property, which exist in other countries.

In principle, the registration tax operates at all levels of government, in accordance with the nature of the act registered. It also competes with a host of estampillas, or stamp taxes, that all tiers can apply on top of the registration tax. The departments are, in particular, entitled to levy—within limits—revenue through stamp taxes for departmental development and rural electrification. Revenue from the stamp taxes is also earmarked. Furthermore, there is a 10 percent surcharge on the registration tax and the local property tax (impuesto predial), tied to the financing of military, agricultural, and cadastral maps. This surcharge has, however, been handed over to the departments, and the surcharge on the property tax was abolished in 1990. All increases of the basic registration tax are also earmarked to finance programs of social assistance.

Given the potentially high base of the registration tax for property transactions, including the separation of property in the case of divorce, and for inheritances and gifts, the revenue from this tax is comparably small—and is only a fraction of the tax on motor vehicles. This is an indicator of administrative complexities and poor enforcement of the tax. It may be worthwhile to simplify the registration tax, in conjunction with that of stamp taxes, and to target the former to the more important economic transactions and legal acts.

Gambling Taxes

Various taxes on gambling (including monopolies) have grown into a very complex system that is difficult to administer. Given the relatively low yield (which may nevertheless be important for some departments), a simpler structure should be sought that is easier to monitor and to administer.

Reforming Departmental Taxation

Existing Excises

A policy change shifting the collection of excises to the producer level could simplify administration and improve revenue collections. This would eliminate problems relating to “internal contraband.” One option is a national excise administration (by DIAN), with the assignment of revenues to departments. Given that departments express little confidence that the national tax administration would be diligent in collecting departmental taxes, DIAN could offer its services for a fee—to promote effectiveness in collection.

Another option is for decentralized excise taxation at the production level, imposed by departments on enterprises within their own jurisdiction. However, this arrangement would favor the producing departments and could exacerbate regional inequities.26 A redistribution mechanism would need to be established to compensate the losing departments.

Excise taxation at the producer level would facilitate tax administration by reducing the number of taxpayers to those with generally well-kept records. Given the production basis of the excise applying to the producer or importer, it would be possible to cross-check the value of total sales with the various departments and the declarations for the VAT. An appropriate margin for the distributor would have to be added in order to approximate distribution based on consumption. Imported excisable goods should face the same treatment as domestically produced goods—with taxation at point of entry—rendering DIAN responsible for collecting the departmental excises on imported goods, in addition to customs duties.

Additional Excises

There is scope for an excise on industrial nonalcoholic beverages administered by the departments at the producer level—with tax collection at the point of origin (importer or producer). This should pose no particularly complex administrative problems for industrially produced drinks. A similar redistribution mechanism could be achieved as for alcoholic beverages.

The Gasoline Tax

At the departmental levels, gasoline tax could be a major source of revenue. Since domestic prices for gasoline are low in Colombia compared with international prices, the scope for gasoline taxation is potentially high. In some European countries, revenues from the gasoline tax rank closely behind the VAT and income taxes. In Colombia, however, the scope of a gasoline excise is limited by a national energy policy. As long as the consumer price of gasoline is administered, any increase in the tax rate would be shifted backward onto suppliers—and would realistically be feasible only if the consumer price of gasoline is allowed to reflect the tax.

Departmental Sharing of Tax Bases with National Level

The sharing of tax bases among governments has to be distinguished from tax sharing. The sharing of tax bases gives the lower tiers direct access to a substantial tax base, such as the income tax, which, for administrative purposes, may be collected by the national administration. Lower level governments are, in principle, free to set their own rates in these cases, which is not possible in the case of tax sharing.

The sharing of the VAT base among levels of government entails severe administrative complexities and regional inequities, owing to the tax credit mechanism and the zero-rating of exports. Such an arrangement would be further complicated if departments were allowed to set their own tax rates.

The simplest surcharge would be for the personal income tax—with the same definition of taxable income as for national taxation. Such a surcharge would benefit most the already better-off departments, given the high correlation with higher income taxpayers, and some of the poorer departments might not benefit as much—thus increasing horizontal fiscal imbalances. The departments would decide on a surtax rate, which may be subject to a national “floor,” to be applied on the national income tax base. The tax rate could also be limited by a ceiling in order to reduce locational distortions resulting from tax-induced migration.27 Initially, there could be a mandatory uniform rate, to allow departments to gain experience with their new tax base, and to register taxpayers. Variable rates could be introduced in a couple of years.

If the total tax burden is to be kept constant, the national government would have to make room for departmental taxation by revising its own rate structure. This would have implications for national expenditures, as well as transfers; the surcharge should not be introduced if these adjustments are not possible.

Revenues generated by the surcharge should accrue on the basis of residence according to the benefit principle of taxation. Withholding taxes should be treated on the same basis. The same taxpayer identifier number is available for the withholding tax as for self-assessment, making allocations to departments possible. Departmental tax collections would not be practical,28 although responsibilities for accounting, auditing, and enforcement could be at the departmental level.

Surtax on Company Income Tax

A surtax on the company income tax, based on the residence criterion for firms, would pose administrative difficulties, given the problems of distinguishing returns at company headquarters from the income’s department of origin. Nevertheless, it is possible to apportion total profits taxable on the basis of simple formulas.29

Environmental Taxes

Ecological taxes are usually linked to negative externalities, to “internalize” the social costs of production and consumption. The price of goods produced and consumed should reflect their “full” costs, not only private costs. Negative externalities may be global, national, departmental, or municipal. Each level of government should have appropriate instruments to contain environmental damage within its territory. Given that the effects may be spread over a number of regions, the charges generally operate within a common framework of national standards. However, administration is often decentralized, also permitting enhanced local revenues.

In order to reduce excessive packaging, a tax could be used to render it more costly, and to finance the disposal of non-recyclable containers. Such a tax could apply to the producers of containers, bottles, cans, cardboard boxes, and similar packaging products. Since the location of such firms is highly arbitrary across the country, the tax would be best administered and collected by the national government. This does not preclude the transfer of the proceeds to those territorial entities that face the daily problems of garbage collection and removal.

There is a scope for further surcharges on utilities, for instance, on telephone bills, electricity bills, or the supply of water. These are easily administered, and generate stable and secure revenues for departments or municipalities. Some of these surcharges relate to local ability to pay (telephone bills), others are broadly connected with environmental damage as caused by the use of energy and water.

Appendix II. The Municipal Tax System

A local tax system should encourage accountability and obey the benefit principle in order to be efficient. In principle, the Colombian property tax (impuesto predial unificado) and the business tax (impuesto de industria y comercio) meet the required criteria. The former can be seen as equivalent to the consumption of local services by local residents and the latter constitutes an equivalent of public services consumed by local business. Furthermore, user charges and fees, as well as betterment levies for the financing of local projects, are particularly suited to the financing of local projects, the benefits of which are more visible and tangible at lower than at higher levels of government.

The Property Tax

The Colombian property tax is essentially ad valorem. In 1990, various taxes on local property were merged into the impuesto predial unificado, which now covers all local properties of households, public establishments, and industrial and commercial enterprises. A part of the tax, 10 percent, is tied to housing construction, and there is a surcharge on the tax (between 15 percent and 25.9 percent), tied to programs for protecting the environment and natural resource bases, 50 percent of which must be spent within the jurisdiction of the local government that collects the surcharge.

The base of the property tax is the cadastral value or, alternatively, a self-assessed value of such property. The latter cannot be lower than the cadastral value. The cadastral value is established officially by the national Instituto Geográfico Agustín Codazzi (IGAC) at regular intervals.30 This institute also provides other services to local governments, such as the dissemination of information and training of officials. A minimum tax base applies, which is calculated in terms of square meters of the land or the building times a minimum price, which varies in accordance with location and social characteristics of the population.

The updating of values is based on information obtained from the register of property transactions and other documentation, which fix reference prices for selected property (predios testigos). These reference prices are used to evaluate neighboring properties across the more homogeneous areas, where various characteristics are taken into account and consolidated on the basis of weights obtained from regression analyses. For intermittent years, the cadastral value is actualized on the basis of the average national consumer price index, as determined by the national Statistical Office (DANE). The National Council for Economic and Social Policy (CONPES) decides on the degree of adjustment to the price deflator, which may range between 70 percent and 100 percent. Special restrictions apply to the adjustment of property values for agricultural land.

In general, property values are significantly undervalued in Colombia and do not reflect market developments. This is explained by various factors: deficiencies in the register of property transactions (the market value is seldom declared); significant time lags in updating the cadastral values; a preponderance of social factors in the evaluation process; limitations on the annual increment of the tax; and concessions expressed in less than full indexing of property values for intermittent years.

Besides the cadastral value, the law allows self-assessment of the property tax value, which cannot be below the former. One would expect that self-assessment would be seldom evoked. There are incentives to induce an automatic adjustment by the taxpayer. For instance, when the property is used as a collateral for a loan, banks are obliged to report the mortgaged value to tax authorities. There are also fiscal incentives that should encourage self-adjustment of property values. The national capital gains tax on realized property transactions acknowledges the registered value of a property for the predial as a fictitious “purchase” price that is deductible from the sale price. By adjusting the registered value for the property tax, the taxpayer can thus reduce his capital gains tax liability. Of course, this incentive works only in the case of intended sale of the property, and it rests on the assumption that transactions prices are declared honestly.

Another problem related to the property tax base is its comprehensiveness. While IGAC claims to cover 97 percent of the properties it is responsible for, the coverage ratio for Bogotá was said to be only in the order of 30 percent. Such figures have to be interpreted with caution, however, since the coverage rate hinges on complete information and timely updating. Every change in the use of property, and every modification of buildings, for instance the partitioning into subunits, may entail a change in the total number of properties. This can only be known after a complete and comprehensive update.

The impuesto predial unificado is the cornerstone of municipal finance and should be strengthened as far as possible. There are essentially two main areas where the property tax needs reform: the valuation of the tax base, and strengthening of tax administration (in particular enforcement of the tax laws). A reform of the rate structure is meaningful once these basic problems have been resolved.

Valuation of Property

The IGAC is the central body in charge of setting the official value of most properties in Colombia (with the exception of Antioquia, Bogotá, Cali, and Medellín). It could be converted into a more flexible and responsive fee-for-service institution that is subject to competition.

A first step could be a decentralization of IGAC, where the 7 regional, 21 sectional, and 44 delegated offices obtain greater autonomy and are allowed to compete with each other. While the IGAC may retain the monopoly to determine officially property values in the country, valuation itself could be performed by municipalities, the regional bodies of the IGAC, and even private institutions (trust funds, banks, and insurance companies). A fee for service could be based on the increment in the value of the property, which would create an incentive to update values on a timely basis. Valuations could be contested by the taxpayer, through reconciliation and legal procedures.

A next step would be a much closer cooperation between IGAC (specifically its autonomous regional bodies) and municipalities. In principle, any municipality can ask IGAC to revise the valuation of any property at any time, based on documented evidence such as the price of a contracted property transaction. In practice, however, most municipalities tend to wait for a more comprehensive revision, stipulated to take place every seven years.

On-line access, or exchange of information, between the IGAC and the municipalities would significantly enhance the quality of the database. Revisions to the value of a representative property should automatically lead to an adjustment of all properties in a given property zone. This would be possible if IGAC procedures for the comprehensive routine revisions based on predios testigos are adhered to. If the cadastre of IGAC is not available on a database, an appropriate project to develop such a system should be initiated. Since the technology for managing the cadastre is sophisticated, and demands highly qualified staff, the cadastre itself should perhaps best be administered centrally.

The evaluation procedures should be subject to competition and debate. At present, the IGAC bases its evaluation mainly on transaction prices which, for various reasons, are often defective. Alternatives are possible, through banks and insurance companies that finance the acquisition of properties or mortgages, based on procedures that include, inter alia, a minimum rate of return, expected rentals, the current costs of construction (in the case of an existing building), and socioeconomic and risk factors.

The national government may also want to reform the legal and procedural framework for property transactions in Colombia, in particular, registration. It is possible to obtain or falsify titles of property, which are used to obtain credits and/or “sell” land. Furthermore, it is common to declare transaction values well below actual market prices. An effective system of penalties would have to be put in place.

A penalty could also be imposed in the form of the government’s right to enter any pending contract at the declared price, plus an appropriate margin. This right would have to be exerted from time to time and should be widely publicized in the media, in order to have a deterrent effect. It may also be prudent to require the payment of all pending taxes relating to property prior to the registration of a change in ownership.

Self-Assessment of Property Values

In 1994, Bogotá introduced self-evaluation of properties, with a remarkable success in terms of revenue raised. Tax collections for 1994 exceeded those of the previous year by roughly 100 percent. This can be explained by a number of factors, the most important of which is the much wider coverage of taxpayers, a consequence of the previously low coverage rate of the cadastre in Bogoté. A further effect relates to previously very low cadastral values with regard to market levels. Such revenue increases may not occur in other municipalities, where the coverage of the cadastre is more comprehensive and where cadastral values have recently been updated. To be fully effective, however, the municipality must exercise its prerogative to purchase property it considers to be undervalued.

The mandatory use of self-assessment has been proposed, but may be premature, particularly given the constraints on tax administration capabilities in many municipalities. Even in Bogotá, a number of administrative issues remain to be solved, and the initial success needs to be sustainable over the medium term. Bogotá, through its policy, has generated a large amount of information that is now unrelated to the cadastre. This creates the need to establish a separate database, requiring new management tools and procedures to keep the new database current.31 Municipalities should therefore remain free to evoke the option of self-assessment whenever they feel to be in a position to handle the flow of information this would entail. If Bogotá can display a well-functioning system in the medium term, the demonstration effect would make a mandate unnecessary.

Rate Structure of the Property Tax

At present, the law allows municipalities to vary property tax rates between 1 and 16 per thousand, which is a relatively wide range. But municipalities usually do not exploit the full scope of potential offered by the law. On the contrary, there is evidence that some municipalities have reduced their rates as property values were reassessed upward. This indicates a lack of volition to use the property tax on the part of some municipalities, where private concerns of local land owners may collide with public interest. The taxable capacity of local governments can also be enhanced by establishing a higher tax rate floor by national legislation. The rate structure of the property tax could also be simplified by eliminating the progressivity according to property use and socioeconomic characteristics of the owner. Such criteria should already be taken care of in a proper evaluation of the property.

A distinction has to be made between rural and urban properties. Rural properties are often small in Colombia, they generate little revenue, and they provide subsistence incomes to poorer families. The strain on tax administration would be considerably reduced by exempting such farms. A marginal adjustment in the tax rate would more than compensate for the reduced number of farms covered. The annual output on the farms greatly exceeds the assessed values per hectare from the cadastre. This suggests that a revision in property values could be mandated on the basis of potential output—greatly increasing revenues, without affecting the poorer farmers. The commercial use of land, for instance, for greenhouses, should not benefit from the exemption.

The Business Tax

The local business tax covers all industrial and commercial activities, as well as services, including banks, savings institutions, financial corporations, and insurance companies. With the exception of Bogotá, the tax base is the average monthly value of gross receipts of the year preceding the tax period. The distributors of petroleum derivatives are taxed only on their distribution margin, which is fixed by the national government. The proceeds from the sale of assets and from exports are exempt from the tax.

The tax rate is set at the discretion of the municipality, and can vary between two and seven per thousand for industrial activities, and between two and ten per thousand for commercial activities and services. There is a surcharge of 15 percent on the amount of business tax collected, which is motivated by the intention to tax advertisements and publicity. This is only a presumptive surcharge since the level of gross sales may not be correlated with this type of activity.

The Colombian business tax has a number of conceptual problems: there is uncertainty with regard to the tax base in the case where a firm produces industrial goods and performs commercial activities (or provides services) at the same time. According to the law, this would give rise to double or even triple taxation. The courts had to clarify that the tax is due only once when production and distribution are carried out by the same firm in the same municipality. Double taxation persists, however, when production and distribution do not take place in the same community.

Of course, double taxation applies for each transaction among different firms, which entails cascading of the tax. This tends to encourage vertical integration of industries, but it is not at all clear how the tax would treat interregional trade within one business conglomerate that operates in various municipalities. The law is silent on the sharing of the tax base in these cases, and the courts may well rule that such transactions be treated like those among independent firms.

Cascading is highly distortive and inefficient, and it favors foreign suppliers over domestic producers and suppliers, in that the former have to pay the tax only once. This bias embedded in the tax base may be tolerable in view of the low tax rates, and given that the total revenue of the tax is low on average (0.4 percent of GDP in 1993), although the tax burden may be heavier on certain sectors. If the tax is to produce higher revenue in the future, however, some reform of the tax base would be desirable.

Options for Reform of the Municipal Tax System: The Business Tax

The local business tax employed in Colombia is essentially a gross turnover tax. Options for the reform of the base of this tax are considered below.

Net Turnover Tax

In order to eliminate the distortive effects of cascading, the business tax could be converted into a net turnover tax. Such a tax is employed, for example, by municipalities in Hungary. The tax base is normally the sales revenue of products sold and services provided, reduced by the purchase value of the goods sold, and the value of services provided by subcontractors.

This option mitigates some of the distortions related to cascading. It is also different from the VAT and it captures both exporters as well as suppliers to the domestic market—in line with the benefit-principle. There are also rules that allow the apportionment of the tax base among different municipalities, if the entrepreneur’s activity extends over more than one locality and if local turnover cannot be defined separately. If this option were to be adopted by Colombian municipalities, it would require the revision of tax rates, since the tax base of a net turnover tax is narrower than that of a gross turnover tax. It is also somewhat more complicated in administrative terms.

Retail Sales Tax (Wholesale Sales Tax)

Cascading could also be avoided if the business tax were levied only at one stage of processing, for instance, at the retail or at the wholesale stage. A retail sales tax (a wholesale sales tax) has a number of problems. It favors municipalities where commerce is predominant, and it disfavors local governments with agriculture, industry, and services, because these activities would remain untaxed. This can lead to severe regional imbalances if the tax is levied at the municipal level. It would also violate the principle of benefit taxation at the local tier, because some business firms would be allowed to consume local services without contributing to their financing.

From a tax administration perspective, a retail sales tax can be far more complicated than the industry and commerce tax. Currently taxpayers report annual sales from the previous year, and pay 112 of that amount each month. With a retail sales tax, taxpayers would have to report sales each month (based on the cash register receipts for audit purposes), and segment products sold, if some are exempt or taxed at different rates. The taxpayers’ compliance burden is further complicated if the sales tax base is different from the VAT base. Given the limited audit capacity of most municipalities, a retail sales tax would be more difficult to administer than the current industry and commerce tax.

Taxation of Factors of Production

Some European countries tax the use of factors of production rather than output. The business tax is considered to form part of the production costs, which include the use of local public services. Often the factors of production are combined in a schedular tariff with a wage and a capital component. This avoids factor distortions.

The payroll taxes on the wage bill are already high in Colombia and were increased recently to cope with new demands for health services and old-age pensions. There is thus limited scope for utilizing this base for local taxation.

Taxation of Local Profits

Another possibility is to tax local business profits. As mentioned earlier, Portugal employs such a tax at the municipal level, the derrama, where the local tax base is derived from global profits within the nation by an apportionment formula. Germany’s business tax also has a profit component (apart from local business capital) that is, however, defined in gross terms. In particular, the deductibility of interest payments is disallowed—which is different from the logic of the company tax.

Presumptive Business Taxation

There are various forms of presumptive business taxation based on objective criteria, such as the number of employees, square meters used, electricity consumption, and the telephone bill. Such forms of taxation are, of course, only approximations to a fully fledged business tax, and their use is mainly motivated by administrative considerations. Since each of these indicators will have a different significance for various business activities, a presumptive tax system has to classify such activities and attribute different weights to the criteria for each class of activities. This introduces a degree of arbitrariness, it creates severe delineation problems in practice (with a potential for conflict and legal disputes), and it complicates the administration of the tax. It should be noted that in Colombia, an unwieldy system of presumptive taxation has already been replaced by a more standardized approach across the whole nation.

A particular form of presumptive taxation could be considered, however. Some municipalities employ a lump-sum minimum tax for small businesses, which acts as a “license tax.” This facilitates tax administration enormously, as it can be controlled even for the occasional street vendor, who would have to carry his “license” issued only after payment of the tax—eventually to be renewed every month.

Local Surcharges

Local surcharges on national or departmental taxes can simplify tax administration, while leaving municipalities their full fiscal autonomy, in particular the discretion to set their own tax rates. Municipalities should avoid application of such surcharges on smaller tax bases, which renders the tax system unwieldy, and could become a nuisance to the taxpayer.

Some municipalities have introduced (or are considering the introduction of) a local surcharge on gasoline sales within their constituency. National laws allow municipal governments to use such a surcharge up to a ceiling of 20 percent. This surcharge, while constituting a potentially important revenue base, could create severe problems of horizontal tax competition, if levied at the municipal level. The problems are similar to those discussed at length for excises. Consumers could evade the surcharge by purchasing gasoline in nontaxing or low-taxing municipalities, where distances are not too great.

In the case of Medellín, the tax is levied on the sales of the main local producer or distributor, not at the local gasoline stations. This creates an incentive for gasoline stations to purchase their supplies outside Medellín, where wholesale distribution of gasoline is untaxed, and they reap an extra margin on their sales in Medellín where the normal retail price would include the tax. Not only is the local surcharge avoided in this case, but the tax wedge is appropriated by private firms at the retail level.32

There are difficulties faced by regional suppliers-distributors with electricity-based surcharges. Such difficulties may be explained partly by administered pricing, partly by the need for heavy investments that amortize only over a longer time period. Again, a more market-oriented approach in the energy sector, combined with targeted social assistance to the poor, is preferable to the subsidizing of production and distribution, but this would require prior adjustments of sectoral policies at the national level.

Slaughter Tax

Given the limited revenue sources for municipalities, the slaughter tax is important at the local level. However, the base is split, with departments being assigned the revenues from the slaughter of large animals, and municipalities getting revenues from the slaughter of smaller animals. This tax should be consolidated at the municipal level.

Appendix III. Revenue-Sharing Indices

Index of Unsatisfied Basic Needs

The index of unsatisfied basic needs is used as a tool for the redistribution of resources from rich to poor areas. This index, widely used for similar purposes in a number of other Latin American countries, is also taken into account for the distribution of cofinancing funds. In Colombia, the index includes the number of people without health and education services, and a few characteristics related to the conditions of housing. While there is clearly no general objection to the use of indicators of need for redistributing resources among levels of government, there are some problems associated with its introduction in the formulas in the Colombian case.

In a diverse country, different factors reflecting need may have more or less significance in various regions (for example, the absence of heating and proper roofing may be more important in the mountainous regions than in the sierra or the coastal areas, where thatched housing without heating would be quite adequate). Moreover, individual jurisdictions have different problems and priorities. Some departments may have adequate provision of access to education and to basic health, and may have other priorities not foreseen in the centrally determined formulas. Since equal weights are given to different components of the index, the formula cannot take into account these diversities and the corresponding needs. The NBI index also excludes other basic needs relevant to the operation of territorial governments, such as infant mortality, unemployment, or the nutritional content of the diet of the poorest strata of the population.

In part, some of the problems may be solved by a reconsideration of the index, based on effective experiences in the various areas of the country. An option is to revise the entire framework of transfers to territorial governments by making this consistent with the decentralization objectives, for example, by making territorial governments more able to respond to local conditions, through an enhanced reliance on general transfers and own revenues at the margin. Specific purpose grants should use indicators (other than the unsatisfied basic needs) better targeted at the specific problems they aim to solve.

Fiscal Effort Index

A fiscal effort index is used for distributing the situado fiscal. This index has a small quantitative impact, but its use may have a severely distorting influence.33 The index is calculated as the increase over the previous year of expenditures on health and education, financed by revenues other than the situado fiscal, that is taxes, fees, profits from locally owned firms, and borrowing. Departments may thus increase their indebtedness, increase their expenditures and obtain a substantial increase in their transfers, even with a reduction in their own tax revenues. Moreover, this index is very sensitive to changes from one year to the other.

Proper fiscal effort indices should measure the actual use by territorial governments of their potential tax base, taking into account differences in the latter across jurisdictions.

Fiscal Efficiency Index

This index is used for distributing the participation transfers to municipalities. Here also, the index has a small quantitative impact. The index is calculated as the rate of change, over two years, of tax collections. This percentage is then weighted by the unsatisfied basic needs index, to favor the poorest jurisdictions. While this index is better than that used for departments (since municipalities have a wider and more standardized tax base than departments), there are still some problems to be solved. Here again, fiscal efficiency is not properly measured. For example, municipalities with a high increase in population, or with new economic activities, may have a huge increase in their tax revenues (and be correspondently rewarded by this index) even if there is reduced effort in tax collections.

Administrative Efficiency Index

The formula for transfers to municipalities uses this index, which is measured as the ratio between per capita operating expenditures and the number of inhabitants with water, sanitary, and sewerage services. It is aimed at stimulating expenditure for these services, since for equal levels of expenditures, it rewards those jurisdictions that have a smaller number of inhabitants serviced. Here again, there is a clear incentive to expand borrowing instead of tax revenues, and to reward jurisdictions with higher degrees of inefficiency in their expenditure.

This chapter is based on developments in Colombian intergovernmental finances through the mid-1990s, and draws on two earlier papers by E. Ahmad, C. Brosio, P. Bernd Spahn, and C. Vehorn and by E. Ahmad, J.-L. Ruiz, and I. Garrido. Comments on these papers were received from Jose Antonio Ocampo, Guillermo Perry, Armando Rodríguez, Emil Sunley, Teresa Ter-Minassian, Andrew Tweedie, and Nestor Urrea.

In this paper, state-level governments are generally referred to as departments (or departamentos, as they are known in Colombia), and local governments are generally referred to as municipalities.

The debate about whether Colombia should have a federal or a unitary structure of government has raged for over a hundred years.

See for example Bird (1984) and Wiesner-Duran (1992). The legal aspects of recent reforms are discussed in Perry and Rodríguez (1991), pp. 65–83.

Includes transfers to the regional and local governments and to the public social security system.

For a description of the benefit principle and the current expenditure assignments (in theory) in Colombia, see Ferreira and Valenzuela (1993).

The departmental tax system and options for its reform are discussed in Appendix I.

The municipal tax system and options for its reform are discussed in Appendix II.

For a detailed account of the evolution of the transfer system to territorial governments in Colombia, see Sanchez and Gutiérrez (1994) and Acosta and Yaker (1994).

Also, the earmarked parts of VAT (on cement), the Cojos de Previsión (Social Security Funds), and the old intendencias and comisarías (other subnational revenues) do not enter the pool.

According to the transition rules established in the 1991 Constitution, local governments could not receive less revenue under the new revenue-sharing system than that which they had received under the previous revenue-sharing system.

The law stipulates that at a minimum, 50 percent of situado transfers for health are to be used to finance primary health care. In 1995 Col$238 billion, equivalent to 50.5 percent of the situado, went to finance primary health services.

According to Law 60, the transfer of responsibilities should be completed by 1997.

Revenues deriving from new taxes are not included, for the first year, in the amount to be distributed to the municipalities, as well as those coming from increases in the existing tax rates and from changes dictated by “economic emergencies.”

Furthermore, a general constraint has been imposed in favor of rural areas. They have the right to a share of the expenditure not less than the share of their population in the total population of the municipality.

Among the most important funds created during the 1980s are the DRI (for rural development), the PNR (destined to rural areas with high incidence of poverty and violence), INURBE (for low-income housing), Fondo de Caminos Vecinales (for rural roads), and FINDETER, a development bank primarily aimed at financing urban infrastructure projects at the local and regional level.

Natural resource taxation exclusively at the local level entails severe risks of misallocation. In practice, a balance has to be struck between the local government’s interests and those of the nation to avoid large regional inequities and to make more efficient use of resources in the longer term. This usually calls for some revenue-sharing arrangement between the local and the national governments—which can take various forms, such as production sharing or royalties, and tests ultimately on political negotiation.

About 90 percent of the National Royalties’ Fund’s resources are from oil production.

Population data for 1993 will be used in the formulas beginning in 1998.

Appendix III contains a detailed description of the main indices used to distribute tax revenue among departments and municipalities.

In principle, prolonged use of private auditors may lead to situations of conflict of interest. Thus, this initial use of private auditors has to include the training of tax staff so that, eventually, they will be able to conduct audits on their own.

Revenue from the liquor tax is assigned to the financing of university and regional hospitals (Decree 1222/86, Article 134), the revenue from the gasoline tax to the financing of roads (Decree 1222/86, Article 151), and 8 percentage points of a 48 percent tax rate on beer to the provision of health services (Decree 1222/86, Article 160).

Although the national Dirección de Impuestos y Aduanas is responsible for collecting customs duties on imported goods, it is not in charge of excise duties that have to be paid to the departments (Article 125 of Decree 1222/86).

Eighty percent of tax collections from the motor vehicle tax are assigned to investment expenditures or to the servicing of debt for road programs.

It is interesting to note that this reasoning applies to excise taxation, but not to royalties on natural resources in Colombia where the producing departments and municipalities are granted almost full access to a tax base the regional distribution of which is highly inequitable. It is also doubtful whether the present system of excise taxation achieves the objective of distributing the proceeds among departments according to consumption. The per capita distribution of the excise on liquor, for instance, shows a highly unequal pattern among departments in 1990, which is, of course, partly explained by variations in consumer behavior. The fact that two of the main producing departments—Antioquia and Cundinamarca—also collect the highest per capita liquor tax might indicate that the tax is mainly distributed according to production even now.

In this case, a ceiling would be preferable, because tax competition is less visible than in the case of excise taxation or surtaxes on the consumption of goods.

This is because of the possible lack of correspondence between the residence of the taxpayer and the location of his sources of income.

California had operated a scheme that identified the state’s share in the worldwide profits of multinational companies on the basis of weighted criteria such as the local wage bill, capital, and turnover. In Germany, where the corporate tax is shared between the federal government and the states, a similar formula is used to apportion a globalized amount of tax paid by firms to regions. Portugal operates a municipal surcharge on the company tax, the derrama.

The department of Antioquia as well as the cities of Bogotá, Medellín, and Cali maintain their own cadasters. By law, the cadasters have to be updated at least every seven years (Law 75/86, Article 174).

This does not mean that the self-assessed values in Bogotá are true market prices. This is partly because, as an initial incentive, Bogotá announced it would be satisfied with values at about 50 percent of the true market price.

Petroleum prices are fixed nationally.

This index has determined the allocation of less than 1 percent of shared revenue, on average, and has thus far been ineffective in encouraging a greater fiscal effort on behalf of the local governments.


    Other Resources Citing This Publication