The impact of inflation on tax revenue is a timely subject in public finance, given the current rates of inflation in both developed and developing countries. Study of the subject has so far obtained no “general” result; the impact of inflation on nominal revenue has been found to depend on institutional aspects of the tax structure and the economy. In the United States and the United Kingdom, for example, the real burden of taxation automatically increases as individuals are forced into higher tax brackets. Three conditions are necessary for inflation to increase tax revenues in this way: (1) earnings need to rise with inflation; (2) the tax structure must be geared to collecting more as earnings increase (that is, it must be progressive); and (3) collection must be prompt. In some developing countries, however, tax systems are generally not progressive, nor is collection efficient, so real tax revenues tend to decline under inflationary conditions. A recent analysis of Argentina’s experience focused on the collection lag between a taxable “event” (for example, earning of income or sale of a commodity) and the time when the tax payment is actually received by the government. If inflation is accelerating, this collection lag results in a decline in real tax revenue.
Between these two examples of inflation increasing and decreasing real tax revenues, a “neutral” tax system might be imagined, in which tax rates would adjust to real incomes and collection would be immediate; in this case, the tax burden would remain constant—insofar as it depended on elements of the tax structure. But experience shows that it cannot be taken for granted that adjusting the tax structure can neutralize the impact of inflation on tax revenue. This argument assumes that tax bases (the type of income, expenditure, production, or asset on which the tax is levied) grow proportionately with inflation. In many countries, however, tax bases may grow systematically at different rates than the general increase in prices. This is particularly true in developing countries, which often respond to excess demand by maintaining an overvalued exchange rate on the one hand and by controlling imports and certain key prices on the other. In countries where foreign trade and taxes on a few products provide the bulk of tax revenues, controlling the prices on these goods can lead to a substantial erosion in the tax base, without reducing excess demand. This article will discuss how inflation can erode the tax base and show how this happened in Ghana during the 1970s. Although Ghana provides a particularly clear example of tax-base erosion, the same pattern is discernible in a wide variety of economies in the developing world.
The mechanism of erosion
In dealing with the question of how inflation affects tax bases in developing countries, it is useful to examine the institutional setting. In a “free-price economy,” prices of all goods and money (that is, the price of money in terms of foreign exchange) change automatically in response to excess demand, and each tax base retains a constant share of nominal gross domestic product (GDP). Many developing economies, however, are “controlled,” in the sense that certain key prices—which also tend to be important in determining certain tax bases—are often set by the authorities and may rise much more slowly than the general price level. Certain sectors of these economies, moreover, typically bear a heavy burden of taxation. Taxes on foreign trade, for example, and sales and excise taxes are important sources of revenue—with the bulk of excise revenue often coming from a few products, such as beverages, tobacco, and fuel. Even corporate income taxes are frequently derived from a few easy-to-tax sectors, such as sugar, coffee, or minerals.
Since the prices that are controlled may well affect tax bases that are important sources of tax revenue in developing countries, such control will reduce the real value of the tax revenues they yield. The thesis that key prices and sectors do not maintain their value in developing countries in inflationary circumstances rests on two propositions. First, that inflation is often accompanied by an overvalued currency and thus by a decline in the foreign trade tax base; and, second, that inflation is accompanied by domestic controls and thus by a decline in the sales and excise tax base.
If a floating exchange rate is adopted, or if a currency is depreciated as domestic inflation increases, then inflation will not cause the foreign trade tax base to decline in real terms. But in many developing countries the authorities are reluctant to depreciate the currency to the full extent of domestic inflation for a variety of reasons. Some of these reasons may be economic. They may believe, for instance, that the supply of export goods will not respond to the incentive of exchange rate depreciation; and they may also believe that an increase in the local price of imports will further intensify inflation. If imported goods weigh heavily in the cost of living of large segments of the population, depreciation of the currency may be perceived as “politically” difficult. It may also be noted that a change in the official exchange rate is a discretionary measure that calls for action by the authorities. For this reason alone, movements in the rate may tend to lag behind the general increase in domestic prices.
Thus, a typical policy response to domestic inflation may be, first, to maintain an overvalued exchange rate. Overvaluation alone would tend to expand the foreign trade tax base by encouraging imports, which are generally more heavily taxed than exports in developing countries. However, given the relative absence of capital inflows characteristic of the developing world, imports are limited by the export trend, which in turn is slowed because the overvalued exchange rate discourages production for export. The authorities will then tend to impose controls on imports, often through a foreign exchange licensing scheme. It is the combined impact of the overvalued rate and import controls that leads to a decline in the foreign tax base, as inflation accelerates and as the authorities continue to manipulate the official price of foreign exchange.
The second proposition concerns the relationship between excise and sales taxes and domestic output or sales. A common response to inflation in developing countries, as elsewhere, is to control the prices of certain necessities—generally in order to control the consumer price index and thereby, it is hoped, to control wages and limit further inflationary increases. As inflation accelerates, price controls are extended to a wide variety of domestic goods. Since excise and sales taxes are typically based on “official” rather than “market” prices, and since the prices of controlled items by definition rise at a lower rate than general inflation, it may be expected that excise and sales tax revenues will also become smaller as a share of GDP.
The primary effects of tax-base erosion in a controlled economy are on foreign trade and sales and excise tax bases. But there are secondary effects: taxes on income and profits may also be eroded as price controls and import licensing alter the distribution of domestic income, shifting incomes earned from “taxable” to “nontaxable” sources.
This process begins as excess demand creates parallel markets in which goods imported at the official exchange rate and scarce foreign currency command more than their official prices. The extent of such parallel markets depends on the degree to which the exchange rate is overvalued, and on the will and ability of the authorities to enforce official prices. The growth of these markets tends to transfer income from those who are obliged to sell at official prices to those who are able to sell at parallel prices. Since price controls are often more effectively administered at the factory or wholesale level, profits of domestic producers may be effectively reduced, particularly those of larger enterprises which tend to be more effectively taxed. (Importers may also be expected to report profits on the basis of official prices.) While the taxable incomes and profits of producers are reduced, middlemen who buy at official prices for resale in parallel markets—and who typically fall outside the limited tax net of developing countries—gain higher income and profits.
Ghana: trade-base erosion
The mechanism of tax-base erosion is widely applicable; many developing economies experience inflation without depreciating their exchange rate proportionately and attempt to suppress excess demand through import and price controls.
As an illustration, Ghana experienced chronic excess demand between 1972 and 1978, which was reflected in an accelerating rate of domestic inflation, fueled by an increase in government expenditure that was financed by bank borrowing. In addition, relative price distortions slowed the growth of exports and placed a constraint on imports, leading to shortages of imported inputs and finished goods. Instead of allowing the price mechanism to absorb excess demand fully at a time of rapid growth in money and credit, as well as shortages on the supply side, the authorities attempted to maintain a system of price controls on domestic goods and a fixed exchange rate, supported by an import licensing system. Thus, while part of the demand for goods and foreign exchange was met at officially controlled prices, as the disequilibrium persisted, the larger part was satisfied in parallel markets.
Given Ghana’s setting of excess demand and inflation in a controlled economy, what happened to government revenue? Estimated data for national accounts between 1969 and 1978 indicate that total revenue as a share of GDP fluctuated but showed a definite downward trend. Revenue reached a high of 20.5 per cent of GDP in fiscal year 1970/71, fell to 15.5 per cent in 1974/75, and dropped more sharply in the period of accelerating inflation after 1974/75. The data show that this fall in the tax ratio was largely due to the slower rate of growth of tax bases tied to controlled prices on domestic goods and import prices set at a fixed official exchange rate.
Since taxes on foreign trade account for 30-40 per cent of government revenue, the foreign trade tax base (imports plus cocoa exports) can be examined first. National accounts data show that imports expressed as a ratio of GDP tended to fall in the period from 1967/68 to 1972/73, to rise in the next two years, and to fall sharply in the years from 1974/75 to 1977/78, when domestic inflation began to accelerate rapidly (see Chart 1). Cocoa bean exports (of the official marketing board)—which represent about 45 per cent of foreign trade in this formulation of the foreign trade tax base—followed a somewhat similar trend, particularly in the sharp fall associated with the last four years of the period. In 1976-77, world cocoa prices rose sharply, and this is reflected in the rise in the foreign trade tax base in the chart.
Chart 1Ghana: rates of change in foreign trade tax base, GDP and consumer price index
Source: Bureau of Statistics, Ministry of Finance and Cocoa Marketing Board, Accra, Ghana.
An alternative method of showing relative movements between the foreign trade tax bases and GDP—since national accounts data may be unreliable—is to compare the annual percentage changes in GDP, in the consumer price index, and in the foreign trade tax base (exports plus imports). This comparison demonstrates that, while GDP and consumer prices tended to rise after 1970/71, the annual changes in revenues from foreign trade were irregular but distinctly lower, particularly after 1974/75.
Since export receipts in local currency, as well as the ability to import, are affected by factors other than the level of the exchange rate, it is not always easy to prove the correlation between an overvalued exchange rate combined with import controls and a relative decline in the foreign trade tax base. However, the data for Ghana demonstrate the link on two fronts: both the foreign trade ratios to GDP and the annual rates of change indicate that foreign trade tax bases have not kept pace with GDP and that the downward trend has been particularly clear in the recent years of accelerating inflation.
Sales, excise erosion
In studying the trends in revenues from sales and excise taxes, an initial difficulty is the lack of reliable production or sales data such as exist for official exports and imports. However, a large proportion of the revenue from excise and sales taxes comes from three categories of goods: beverages, tobacco, and petroleum products. In addition, excise and sales taxes are levied at the factory stage, where prices on tobacco and beverages are effectively controlled. Petroleum products are produced by a state monopoly that sets internal prices on which taxes are levied. Thus, the wholesale price index for beverages and tobacco and for fuel and electricity is used as a proxy for the tax base on these products. The assumption of a constant level of real output is probably realistic because shortages of imported inputs led to a stagnation in production over the period as a whole. Using these indices as proxy tax bases, the rates of increase in wholesale prices for beverages and tobacco and for fuel and electricity can be compared with the rate of increase in the consumer price index and in GDP (see Chart 2).
Chart 2Ghana: increases in prices of excise products, GDP, and consumer price index
Source: Bureau of Statistics, Ministry of Finance, Accra, Ghana.
For the period 1968/69 through 1973/74, the controlled prices for the major excise products generally kept pace with increases in average consumer prices and GDP. But in 1975, as inflation continued to accelerate, increases in the prices of excise products began to lag behind. As in the case of imports and exports, the excise tax base failed to keep pace with inflation in the rest of the economy.
In addition to its impact on total government revenue through the erosion of tax bases, a dual price system, such as the one in effect in Ghana, may also have a powerful impact on the allocation of resources. As noted above, the primary impact of the maintenance of an overvalued exchange rate combined with import controls is to reduce the foreign trade sector relative to the rest of the economy. Domestic production as a whole may then be reduced because of its dependence on imported inputs.
In Ghana, the effect of the dual price system that resulted from official price controls was felt most strongly in the cocoa and manufacturing sectors. Ghana has traditionally supplied about one fourth of the world’s cocoa production; cocoa is the foremost source of foreign exchange and a major source of tax revenue. The dual price system means that a farmer can sell his cocoa at the official producer price to the Cocoa Marketing Board or at an unofficial price to the neighboring countries of Ivory Coast and Togo.
How different were these prices? Trends in Ghana’s official producer price may be measured by deflating this nominal price by the consumer price index to obtain a “real” producer price. The producer price may then be compared with that of neighboring countries, using both the official and the parallel market rates for the currency. Using the first measure, a downward trend in the real producer price in recent years was apparent, reflecting a similar trend in farmers’ real incomes. The official producer price fell by about 44 per cent in real terms between 1967/68 and 1977/78. A comparison of Ghana’s producer price converted at both official and parallel market rates shows that Ghana’s producer price in 1976 was below those of its neighbors by both measures. Increases in Ghana’s producer price in September 1978 raised it above those of both neighboring countries when expressed at the official rate of exchange. At the parallel market rate of exchange, however, the producer price in neighboring countries was much higher than the price in Ghana.
Given the low official (producer) price, the cocoa farmer could in theory respond by devoting less effort to cocoa crops, switching to other crops, or smuggling cocoa to neighboring countries. There is evidence that all three effects took place. Official sales to the Cocoa Marketing Board fell from an average of 430,000 metric tons in the 1969-72 period to 277,000 metric tons in 1977/78. Prices of other crops, such as rice and maize, were allowed to rise freely, and farmers seem to have substituted these crops for cocoa. The output of maize, for example, was estimated to have risen from 286,000 metric tons in 1976 to 408,000 metric tons in 1978. An important aspect of the shift for public finances is that cocoa is heavily taxed, while maize is completely untaxed. The amount of cocoa smuggled into neighboring countries is difficult to quantify, but smuggling in 1977/78 was estimated to have amounted to a minimum of 10,000 metric tons.
A less dramatic impact of the dual price system occurred in the manufacturing sector. Since ex factory prices were more effectively controlled than retail prices, profits of farmers were held down, and production and investment tended to decline. As in the case of agriculture, output shifted from a relatively highly taxed activity (manufacturing) to a lower taxed activity (retailing).
Tax-base erosion occurs in a wide variety of conditions in developing countries trying to cope with excess demand pressures by using similar measures. In too many cases, the resulting tax-base erosion may exacerbate the problem of containing excess demand by contributing to a vicious circle in which the fall in real tax revenue tends to widen the fiscal deficit, itself a major cause of disequilibrium. Policy changes intended to forestall the erosion of the tax base that would otherwise occur can therefore make an indispensable contribution to the stabilization efforts of such countries.