DIAL

The primary changes which developing countries have been required to make to their fiscal policies involve reducing revenue generated by trade and increasing revenue generated domestically (This article is a summary of the presentation given bu D. Cogneau at the "Journée des Economistes de l'ORSTOM 1996".). This reform involves revising the rate structure, eventually expanding the domestic tax base, and finally improving the overall efficiency of the tax system (i.e. combating fraud). As it occurs in the context of structural adjustment programmes, we expect it to yield increased economic efficiency, especially owing to an improved allocation of resources and international specialisation. Furthermore, providing that these changes are successfully implemented, an improvement in the government's fiscal position resulting from increased revenues will permit an increase in productive public expenditure. The goal here is not income redistribution, as in the case of tax reforms undertaken in industrialised countries in which the very structure of domestic taxation (income taxes, value added taxes, etc.) is re-evaluated. In most cases, the fall in revenues from foreign trade is compensated by an increase in indirect revenues from domestic production and consumption, notably value added taxes. Nevertheless, this fiscal-tariff transition has important implications for the distribution of income, which vary depending on the nature of both the reform and the economies and societies to which it is applied.

The author of this article hopes to make some contributions of both a theoretical and an empirical nature, drawing on international economics, the dependent economy model, and welfare economics, though it is recognised that this treatment is far from exhaustive. First, we shall examine the budgetary and financial consequences of fiscal and tariff reform, and then their impact on prices and income distribution. Two case studies which were modelled under general equilibrium, dealing with the income distribution implications of the UDEAC (UDEAC is the French initials for the Central Customs and Economic Union.) reforms in Cameroon, and the consequences of trade liberalisation in Morocco, are presented in a box.

I. The Fiscal-Tariff Reform Transition and its Budgetary and Financial Consequences

Budgetary Consequences

We define the fiscal-tariff transition as a decline in tariffs which is either totally or partially compensated, in terms of government revenues, by an increase in indirect taxes (VAT or other) on products (including, of course, imports). This total or partial compensation is particularly justified in countries which have already experienced a drastic adjustment of government expenditures, and where it may be said that these expenditures have attained a floor.

Ceteris paribus, the extent of the required compensating adjustment to indirect taxes is obviously inversely related to the size of this tax base relative to the import sector. An economy with a very large informal sector will see the formal sector absorb a disproportionately high tax increase. Under these circumstances the reform faces a serious risk of failure, as the formal economy, subject to an exorbitant tax rate and facing stiff competition from imports and from the informal sector, may well disappear into the latter. An increase in fraud, a decrease in tax revenues and a budgetary crisis may all follow, so that tariffs retain their dominant position in the government's budget (this was the case in Madagascar and, to a lesser extent, in Cameroon during the reform of the UDEAC zone in 1994). Re-establishing budgetary balance inevitably necessitates adjustments on the expenditure side. In this case, only by extending the tax base to the informal sector can the transition be realised efficiently. The cost of broadening this reach is, however, difficult to estimate, and its distributive impact is probably quite inequitable, as it consists of taxing products (food and the output of cottage industries) consumed by the poorest members of society.

Changes in the Real Exchange Rate

Economic agents consume three types of goods: imports, products of the domestic formal sector, and the output of the domestic informal sector. Prices in these three categories are affected differently by fiscal reform : imported goods are subject to tariffs as well as to sales taxes, goods produced in the formal sector are subject only to the latter, while informal sector output is exempt of all taxes. If the decrease in tariffs is sufficiently pronounced, and if domestically produced goods are only imperfect substitutes for imports, then the reform will result in consumers facing import prices which are lower relative to those of formal and informal domestic goods (We abstract from the issue of smuggling, assuming that legal imports are marketed by the formal sector. In other words, we assume that contraband is neither subject to tariffs nor taxes, while legal imports are subject to domestic taxation.).

Ex ante, we see that this transition has an impact equivalent to an appreciation of the real exchange rate. This is why it is often accompanied by a compensating devaluation of the currency (for example, trade reforms in the customs union undertaken in 1994) to re-establish long-run equilibrium. Often, even in the framework of regional free-trade agreements (NAFTA, the EU-Morocco and EU-Tunisia trade agreements), the increased penetration of imports is much greater than the growth of exports, owing to the asymmetric nature of trade liberalisation tariff barriers being much higher in the South than in the North. Since a primary goal of the transition is to attract foreign private investment (direct investment), however, it should, if it is successful, eventually result in a stabilisation of the real exchange rate. An influx of foreign public capital (aid or loans) may also accompany the reform and serve to temporarily balance the foreign account (These financial flows are, however, compensated in the long-run by flows in the opposite direction repatriated profits and debt repayment.).

The Danger of Macroeconomic Destabilisation and the Political Economy of Transition

It is of some importance not to minimise the potentially destabilising impact of tariff reform, especially when the macroeconomic management capabilities of the government are weak and when significant productivity gains in the tradable goods sectors are not envisaged (required for increased competitiveness and attracting foreign direct investment). This danger of destabilisation is all the greater when the microeconomic efficiency improvements and growth resulting from liberalisation will be minimal.

The recent example of Mexico speaks volumes on this matter persistent balance of payments problems combined with a lack of confidence from outside investors to precipitate a crisis and force drastic adjustments. Mexico, being a large country, is not particularly vulnerable in matters of external debt and budgets, as its openness to the outside is relatively limited and its domestic fiscal situation is more developed than those in the countries of the Maghreb or sub-Saharan Africa.

Rodrik (1996) reminds us to what extent the trade liberalisation of the countries of East Asia and South-east Asia was a gradual process, and that this was preceded by measures to reinforce governments' macroeconomic management capabilities and by a rapid accumulation of factors of production (physical and human capital). In contrast, some Latin American countries have had to accomplish in five years what Asian countries did in over twenty years. It is of particular interest that Rodrik attributes the better "governance" and management capacity of Asian countries to an initially more egalitarian distribution of wealth (which, in addition, creates conditions favourable to a rapid accumulation of human capital).

Let us examine in more detail the gains, and their distribution, in microeconomic efficiency attributable to tax and tariff reform in order to assess how it impacted on income distribution.

II. Expected Price and Income Effects

Income Redistribution in the Traditional Framework

Standard general equilibrium models usually provide poor estimates as to long-term microeconomic efficiency gains generated by trade liberalisation, on the order of two to three percent of GDP on average(Brown et al. (1994), Golding et al. (1993), Robinson (1991) and Rutherford et al. (1995). For the Maghreb and sub-Saharan Africa some estimates are lower, even negative (cf. Goldin et al. (1993) for the impact of the Uruguay round on these regions, for example).).

Traditional economic theory predicts that trade liberalisation will stimulate specialisation in sectors using factors of production which are relatively abundant in the case of developing countries these are natural resources and unskilled labour. The growth in productivity of these inputs induced by the migration of mobile factors into these sectors allows their remuneration to increase relative to that of capital and skilled labour. This line of reasoning suggests that the distribution of income should improve as a result of trade liberalisation as unskilled labour and peasants benefit. However, the case of land is particular. When agricultural imports from developed countries compete head-on with local produce, the ensuing fall in prices hurts land-owners while benefiting non-producing consumers. Incomes of individuals living only on rental income falls. Thus, even if the increased specialisation of the economy in areas of comparative advantage benefits small farmers, the final effect on their incomes of liberalisation is indeterminate, and may be negative. Within this standard framework, the evolution of the income of poor peasants far from the market appears to form the crux of the distribution issue. If this seems to be a burning issue in Mexico (corn growers, peasants in the Chiapas), how much more might this be a concern in the Maghreb or in sub-Saharan Africa.

Redistribution over Prices in the Traditional Framework

On the other hand, these effects on relative incomes may be compensated by price effects. The various classes of consumers (capital owners, skilled labour, unskilled labour, rural, etc.) do not all share the same consumption bundle and hence have different cost of living indicators. The wealthiest citizens, whose bundles contain more imports, tend to benefit from the tariff reduction. However, the compensating increase in indirect taxes, applying to the formal goods, doubtlessly affects the same groups. Furthermore, the fall in the price of food imports may benefit the poorest (cf. the case of Cameroon). The governing factors here will be, on the one hand, the distribution in consumption of imported goods affected by the reform and of the products of the formal sectors (which constitute the base of indirect taxes) in broad terms, their relative income-elasticities and, on the other hand, the substitution elasticities between these two types of goods. At one extreme, when formal sector products are very similar to imports (a high degree of substitutability), their prices fall in tandem with those of imports. Consumption bundles comprising a large share of these goods clearly benefit. However, in this case it is difficult to realise the anticipated budgetary compensation. At the other extreme, if imported consumption goods are complements to formal goods, the increase in price of these latter will counteract the fall in price of imports in the consumer's bundle. If the shrinking of the domestic tax base is sufficiently small and the increase in the rate sufficiently big, the final impact of the reform on the redistribution of the cost of living may prove to be equitable (cf. the case of Morocco).

All things considered, the income redistribution resulting from relative price and cost of living movements should be limited in the majority of cases. Estimates for Cameroon and Morocco confirm the dominance of income effects over price effects.

The pattern we have been examining associates a major redistribution of wealth with a fairly modest increase in overall efficiency. Rodrik (1996) coined the term "political cost-benefit ratio" to describe this concept. This far-reaching income redistribution is further associated with an extensive sectoral reallocation of factors (movement of specialisation).

Naturally, there exist arguments in favour of free-trade which are not based on the standard models, and which vaunt its distributive virtues :

Sectors with a comparative advantage may show increasing returns to scale such that the opening to international trade yields access to huge markets and spurs much more significant productivity, and hence wage, gains.
The monopoly or oligopoly power of previously protected domestic producers may be reduced in the markets of tradable goods, limiting the earnings of capital owners.
The output of exportable commodities may benefit from the increased availability of imported inputs or from the growth in exports. If these externalities allow for capital to be saved and decrease its relative remuneration, we may have a desirable redistribution alongside an efficiency gain.
Trade liberalisation may encourage positive expectations in foreign investors, resulting in a decline in the risk premium on capital inflows. The accumulation of foreign capital will tend to drive down profit rates and the return to capital, and may yield benefits for skilled labour, a complement to capital, thus benefiting the well-off rather than the poor.

Even if the benefits of trade liberalisation include, among others, simplicity and prudence, its effectiveness remains however the subject of debate. The strongest argument against surely arises from models of endogenous growth. Nothing guarantees that the sectors of comparative advantage will be the growth sectors, benefiting, for example, from technological externalities or learning by doing. If these two sectors do not coincide, then there exists an optimal system of tariffs, as the "new international economics" has shown. When trade liberalisation diverts resources (capital, labour, etc.) from growth sectors, efficiency gains and redistribution may occur at the expense of long-run growth. Short-term increases in welfare may then forfeit long-term welfare growth. Current trends in direct investment from the North into the South appear to target technological sectors or sectors with dynamic domestic markets (high income-elasticity, for example), and are concentrated in countries with substantial human resources and expanding markets (primarily Asia). Consequently, it may be dangerous for developing countries to excessively promote specialisation in the traditional sense. From a distributive perspective, at least, the growth sectors named above are probably less intense in labour, or in other resources which are fairly equally distributed, than some of the sectors of comparative advantage. Of course, it may also be deemed inequitable to indefinitely maintain a system benefiting a handful of sheltered firms and their employees if accumulation and growth are not forthcoming.

III. Conclusion

For the tax and tariff reforms undertaken by most African countries to succeed without causing a major macroeconomic destabilisation, it is not only necessary that the domestic tax base be enlarged and that the governments have solid macroeconomic management capabilities, but also that these reforms be accompanied by productivity gains and a significant pace of factor accumulation.

As a rule, the estimates of the long-term, microeconomic efficiency gains of free trade derived from general equilibrium models are rather modest. Trade liberalisation should however be accompanied by a significant redistribution of income. In Africa, the core of the distribution problem apparently concerns the small food farmers, for whom the impact of the fiscal and tariff transition remains uncertain.

Empirical General Equilibrium Analysis using a Micro-Macro Model

The models used here to analyse the tax and tariff reform transitions in Cameroon and Morocco are based on an maquette from Bourguignon, Branson and Melo (1991), designed specifically to study distribution effects in developing countries implementing stabilisation and structural adjustment policies. This dynamic framework accounts for both short-term macroeconomic stabilisation issues through the inclusion of a monetary and financial module, and for longer-term structural phenomena such as reallocation and growth. We will not discuss the characteristics of these models here, except to recall that they account for some rigidities in the economy: rigidity in the technology of production, imperfect substitutability between imported and domestic goods, rigidity in the formal-sector labour market, urban unemployment and the existence of segmented markets connected by migration and limited socio-professional mobility, and the oligopolistic character of formal enterprises (industry and services).

The Tax and Tariff Reform of the UDEAC Zone in Cameroon: Marginally Negative Distribution Effects

The tax and tariff reform undertaken in 1994 and 1995 created the following nominal tax rates: a 35% decrease in import duties on agricultural produce, industrial output and on complementary imports, combined with indirect tax increases on wood and industrial output (+50%) on building and public works (+33%) and on services (+20%). Assume that tax fraud does not increase owing to the tax increases, and that formal companies do not lose market power because they adjust by cutting their utilization of production capacity and labour. We then observe that the transition does not appear to erode the return to capital or wages in this sector, but that unemployment increases. The slight reduction in the budgetary deficit resulting from the reform reduces the crowding-out effect on private investment and increases the demand for building and public works. This particularly benefits non-agricultural rural households living in typical small towns as well as potential migrants into the two main cities. Finally, competition in the foodstuff markets depresses their prices and the income of peasants. This fall in prices of agricultural produce has two consequences: a fall in the cost of living of the poorest, but also a decline in incomes for peasants. The increase in agricultural exports induced by the reallocation of agricultural labour is not sufficient to compensate for the lower revenues, and real incomes of farmers decline. On the other hand, the lower price of food benefits labour in the informal sector of small towns (rural non-agricultural). Both from the perspective of efficiency gains as well as for income distribution, the consequences of the reform appear slightly negative.

Trade Liberalisation in Morocco: A Substantial Decline in Social Inequality Without a Marked Reduction in Poverty

In the case of Morocco, we examine the hypothesis of the extreme case of the total abolition of tariffs (complete free-trade). Notice that this does not actually reflect the agreements recently signed with the European Union, which apply neither to agriculture nor to the rest of the world. In a sense, our hypothesis combines the consequences of this agreement, extended to agriculture and non-European trading partners (who represent one third of Morocco's foreign trade by destination) with those of Morocco's adhesion to the WTO (the Uruguay round was concluded in Marrakech).

The distributive consequences of the trade liberalisation appear considerably more positive than was the case in Cameroon. The greater flexibility of the Moroccan economy allowed sectoral reallocation to play its distributive role. We observe a significant decline in inequality, but a lesser decrease in the poverty indices. Groups whose income depends upon agriculture benefit the most. However, large-scale farmers' experience the greatest rise in incomes, as the increased demand for agricultural exports drives up the return to land. As to agricultural wages, reverse migration from the cities to the countries has a moderating effect. Modern workers, more skilled and better paid initially, suffer from the decline of branches of industry which do not have a comparative advantage. Capitalists, the principal owners of formal capital in these branches, sustain an even greater loss of income. Finally, reverse migration and a more sustained demand for building and public works for investment purposes drive up the salaries of urban informal labour. Cost of living variations for the different groups also contribute to the redistribution of welfare, imports being more broadly spread among the consumption bundles and the VAT affecting primarily the urban groups. In sum, the transition seems to have beneficial distributional consequences in terms of equity. Notice, however, that among the two poorest groups in the population, informal urban labour and agricultural wage-earners, only the former experienced a significant improvement in their living conditions.

Bibliography

Bourguignon F., de Melo J., Branson W.B. (1989), Macroeconomic Adjustment and Income Distribution : a Macro-micro Simulation Model, OECD Development Center, Technical Paper n°1, Paris.
Bourguignon F., de Melo J., Morrisson C. (eds) (1991), Adjustment with Growth and Equity, World Development, Special Issue, vol.19, 11, pp. 1483-1651.
Brown, D. K. and al (1994), An Assessment of Extending NAFTA to Other Major Trading Countries in South America, mimeo, Research Consortium on North American Trade and Investment (3rd annual meeting), Toronto, 21 p.
Cogneau D., Herrera J., Roubaud F. (1996), La dévaluation du F CFA au Cameroun, bilan et perspectives, Economies et Sociétés, coll. " Relations Economiques internationales ", 1, pp. 169-205.
Cogneau D., Tapinos G.(1995), Libre-échange, répartition du revenu, et migrations au Maroc, Revue d'Economie du Développement, 1.
Cogneau D., Razafindrakoto M., Roubaud F. (1996), Le secteur informel urbain et l'ajustement budgétaire au Cameroun : faits empiriques et modélisation macro-économique, 1995. A paraître en 1996 dans la Revue d'Economie du Développement.
Cogneau D., Dumont J-C., Izzo P. (1996), Regional integration, direct investment and migration in the European-Mediterranean area results obtained from a computable general equilibrium model. 32 p., paper presented at the OECD seminar on "Migration, free trade and regional integration in the Mediterranean basin", Athens, 31 October - 1 November 1996.
Goldin I. and al (1993), Trade Liberalization : Global Economic Implications, OCDE-World Bank, 217 p.
Robinson Sh. and al (1991), Agricultural Policies and Migration in a US-Mexico Free Trade Area : A Computable General Equilibrium Analysis, California Agricultural Experiment Station, Gianinni Fondation of agricultural economics, Working Paper n°617.
Rodrik D. (1996), Understanding Economic Policy Reform, Journal of Economic Literature, vol. XXXIV, pp. 9-41.
RutherfordTh.F., Rutström E.E., Tarr D. (1994), L'accord de libre-échange entre le Maroc et la CEE : une évaluation quantitative Revue d'Economie du Développement, 2, pp. 97-133.