JORGE BRAGA de MACEDO

1 Introduction

Integration with the rest of Europe has become quite evident in Portuguese society only four years after the country joined the European Community. The government regards the deadline for completing the internal market by the end of 1992 as a major political challenge. This stance departs from a tradition of ambiguous policy towards European integration, which has shared some features with the British position. Once before, in 1975, there was an attempt to break with this tradition, but on that occasion the new departure was in the opposite direction, when the revolutionary leaders sought to reject the West in favor of central planning along the lines of Eastern Europe. Although it was soon aborted, the attempt had the effect of keeping the public sector frozen until 1989, which muted the economic effects of accession to the European Community in 1986. The market is still heavily influenced by the state, and economic liberalization lacks domestic political credentials. This applies to Portugal even more than to historically dirigiste countries, such as, until recently, France. Yet Portugal's case may be of far-reaching interest. The experience of a country that is undergoing transition from widespread nationalization to privatization within a democratic set-up may be useful to countries in Eastern Europe striving to achieve political democracy and to reform their economies at the same time.

In this paper the tradition of ambiguous response of the Portuguese government is explained as revenue seeking by the state. The paper views administrative regulation in financial markets as implicit taxation and argues that the confusion that has resulted from this disguise has not served Portugal well. Regulation in product and labor markets has had comparable effects. It has resulted in a dual structure, comprising competitive and rigid segments, which generates efficiency losses. This dualism is of long standing. In the 1960s, trade liberalization uncovered previously ignored export potential in manufactures, but at the same time industrial and financial regulation produced conglomerates whose activities spread to the African colonies: they were soon to become state-owned and uncompetitive.

After the revolution of April 1974 the public sector was enlarged through widespread nationalizations and then frozen by an article of the new constitution that made these nationalizations irreversible. At the same time state intervention in the rest of the economy was extended. In some areas, such as the labor market and some parts of industry, the private sector has circumvented or overcome intervention. In recent years the government has begun to dismantle some of the measures taken in the early post-revolutionary years, for example, by providing for greater flexibility in the labor market, by allowing new banks to enter the financial sector and by initiating the privatization of some of the nationalized banks. Privatization necessitated a constitutional amendment to repeal the irreversibility of the nationalizations. The way is now open to unfreezing the public sector.

Much of the impetus to free the economy has been generated by Portugal's increasing integration with the European Community. With accession to the Community in 1986 Portugal entered a period of transition towards full integration with the Community by 1996. The pressures this has generated for liberalizing the economy have since been reinforced by the agreement of the twelve member states to remove all remaining barriers to trade and to factor mobility by the end of 1992. Yet much has still to be accomplished if Portugal is, in the words of the government's slogan, 'to win 1992'. Unless state intervention is reduced, Portuguese firms will have great trouble preparing for the single market. Furthermore, if the economy fails to catch up with the rest of the Community, the government may feel unable to make the commitment to Economic and Monetary Union.

Help from the Community is coming in the form of increased funds for structural programs. The structural funds will reach 4.5% of GDP in 1993, from 2.7% in 1989 and will provide for about one-third of investment over the period. Programs aimed at providing the country with an adequate infrastructure and training are included in the Regional Development Plan (PDR) of 1989. There are also specific programs aimed at modernizing agriculture and industry, respectively the PEDAP of 1986 and the PEDIP of 1989. The Community may also delay until 1996 the deadline for Portugal's financial integration, which will entail the freezing of all capital movements. This paper argues that a precondition for the success of European integration is transparent fiscal policy. Even if Portugal's reported budget deficits continue to fall, hidden deficits may persist - and the tradition of ambiguous public response will be perpetuated.

The remainder of the paper is in six sections. To set the stage, Section 2 establishes the tradition of ambiguous response. It asserts that ambiguity in Portuguese policy towards European integration is not simply the result of the absence of political democracy from 1926 to 1974. In particular, the country's neutral stance during World War II and the enduring colonial ties with Africa made it as averse as Great Britain to the federalist enthusiasm of the late 1940s. A combination of external shocks and domestic policies steered Portugal into a process of catching up with the European Community, which is still continuing.

Next some economic features relevant to the development process are discussed. Employment, emigration and the causes, and effects of real wage flexibility are taken up in Section 3. Section 4 presents evidence of the trade effects of accession to the European Community. The ease with which industry has adapted since 1986 is noted. Simulation of some of the trade effects of the single market program shows that the full impact of external competition has not yet been felt, either on traditional labor-intensive exports or on prospects for intra-industry trade. The role of foreign direct investment and of more open financial markets is stressed.

The following sections attempt to reveal the root of the ambiguous public response to external liberalization. Revenue-seeking regulation of industry and finance is found side by side with trade liberalization. Section 5 points to the trade-off between regulation and competition in banking: it confirms that the propensity towards rationing schemes and the process of European integration have been equally durable. In Section 6 attention is turned to the constraints on fiscal adjustment implied by a public sector which includes banks and was frozen by constitutional law between 1976 and 1989.

Section 7 shows how a frozen banking sector and a disguised fiscal policy constrain monetary and exchange-rate policy. Trends in competitiveness are assessed, and the growing ineffectiveness of the crawling peg regime introduced in 1977 is documented. Displaying the intention of joining the European Monetary System (EMS), as the government has done, is not enough. A credible peg requires convergence of nominal values with those of the EMS partners. But this necessitates disinflation, which cannot last without some restriction of demand. This in turn is predicated on the credibility of fiscal adjustment.

The sequencing discussed in Section 8 with the objective of EMU in mind requires a program implemented over several years, much like the Program to Correct External Imbalances and Unemployment (known as PCEDED) approved in 1987 and revised in 1989 to incorporate the effect of Community structural funds. But a multi-annual fiscal adjustment strategy - which may be called MAFAS - appropriate to 1992 needs to be more ambitious. The share of public expenditure in GDP must be reduced, and taxes must be made more transparent and less susceptible to avoidance and evasion. The paper concludes by suggesting that such a medium-term orientation be made evident in the government's handling of the budget for 1990, which remains too high a percentage of GDP (8%, the same deficit as the initial 1989 budget).

2 Domestic constraints on integration policy

Accession to the EC in 1986 and preparation for the completion of the internal market have provided a renewed opportunity for and impulse towards modernization, which entails liberalizing those aspects of the economy where intervention remains excessive. The government has identified meeting the 1992 deadline as the major challenge facing the nation. The objective of EC economic and social cohesion is seen as requiring Portugal to meet this challenge. Partly as a result of government pronouncements and advertising efforts, the business community is becoming increasingly aware of the need to compete with firms elsewhere in Europe after 1992. Strategies are mapped out with this objective in mind, and fears of failure are prompting a surge of lobbying aimed at delaying integration or at limiting its scope.

Such apprehension exists even though the first years of membership of the EC have been a major success in terms of export expansion and the response of firms, as well as macroeconomic performance. Doubts as to whether these advances can be maintained reflect concern that public sector reform has been far from adequate. The introduction of income tax, the partial privatization of public enterprises and other 'structural' reforms have been too slow to provide the underpinnings for further progress. Furthermore, the inefficiency and uncompetitiveness of the large public sector have made the challenge of preparing for 1992 that much greater for the economy as a whole. Nevertheless, that there is now less ambiguity towards integration is a major departure from previous experience of external liberalization.

The ambiguous response to the challenge of European integration dates back to the 1940s. Then Portugal accepted Marshall aid and so took up the challenge of catching up with the rest of Europe. Yet at the same time the government was determined to hold on to the African possessions, which were turned from colonies into provinces in the 1951 revision of the Constitution. The idea of 'fortress Portugal' was accentuated in the 1960s. In 1974 the revolution maintained the fortress mentality but the domestic objective constraining European integration changed: colonialism was replaced by socialism. Both were understood to be strictly domestic objectives rather than manifestations of international trends.

Before and after the revolution the economy displayed a dual structure with elements both of liberalization and of intervention. In the early post-war period control over public spending and the accumulation of foreign exchange were the keys to defending the currency and securing the financial independence of the state. This implied strong supervision by the central bank. Barriers to foreign competition were part of a regulatory environment that also featured investment licensing in many industrial sectors. Economic exchanges with the colonies were also regulated despite the fortress idea, expressed as an objective of 'national economic integration'. Yet competition prevailed, fortunately, in sectors where in the late 1950s trade was liberalized in the framework of the British-led European Free Trade Association (EFTA). Export growth was so pronounced in clothing that Portugal was sometimes labeled as a 'pajama republic'.

Thanks to sizeable migrants' remittances, tourism and growing foreign investment; gold reserves piled up during the 1960s. Against the background of the first enlargement of the Community to include former EFTA partners Denmark, Ireland and the United Kingdom, Portugal signed a free trade agreement with the EC in 1972. This crowned a successful decade of export-led growth, which was followed by a stock market boom. Existing exchange controls were unable to prevent growing financial interdependence¹.

The rate at which the Portuguese economy has been catching up with the EC over the past two decades can be seen in Figures 9.1 and 9.2. Figure 9.1 plots two ratios of Portugal to EC12 variables from 1970 to 1988: on the left hand scale, gross domestic product per capita at current prices and purchasing power standards; on the right hand scale, the share of gross fixed capital formation in gross domestic product. Figure 9.2 plots the incremental output-capital ratio, given by the growth rate of output divided by the output share of gross fixed capital formation; the variable provides a rough estimate of trends in the marginal efficiency of capital in Portugal during the same period. The figure suggest that the main reason underlying the catching up was either that investment was a significantly larger share of output in Portugal than the average of the 12 EC countries or that it was much more productive.

As Figure 9.1 shows, after catching up rapidly with the European Community in the early 1970s, the Portuguese economy suffered a relative decline. In late 1973 the first oil shock hit Portugal with particular severity. Moreover, the political situation disintegrated to the point where the fifty-year-old authoritarian regime fell without resistance on 25 April 1974. A military junta took power and vowed to 'democratize, decolonize and develop' in the name of the people. Democracy and development soon became equated with an 'original' brand of socialism, but the originality was constrained by the combined influence of the armed forces and of the Communists. The Common Market was seen as a capitalist and imperialist threat. Economic solidarity with the eastern bloc and the Third World was proposed as a superior development strategy.

On 11 March 1975, under the threat of a military take-over, firms in protected sectors, especially the financial conglomerates, were nationalized without compensation to shareholders². To ensure the transition to socialism, an article of the political constitution passed in 1976 froze the post-revolutionary nationalizations as 'irreversible conquests of the working classes' until the second amendment, voted in 1989 overturned this irreversibility. Nationalization without compensation had the effect of greatly enlarging the public sector virtually overnight with no side effects on revenue, while the irreversibility of the nationalizations basically froze the public sector. This seriously hindered the process of economic restructuring called for by the global shocks of the 1970s.

It is evident from Figure 9.1 that the Portuguese economy was already beginning to catch up again in 1977, when the application for EC membership was lodged and agreement was reached on the first stabilization plan involving the International Monetary Fund (IMF). It is worth noting that the combined expectation of trade liberalization and of microeconomic stabilization did not appear to hamper the process of catching up. Flexibility in the economy was such that exporting firms and workers adjusted sufficiently to compensate for the downward rigidity of public expenditure.

The relentless expansion of public debt that resulted from the enlarged public sector made private firms the residual borrowers from credit enterprises. The private sector was subject to recurrent squeezes because of stop-go macroeconomic policies. These were in part politically motivated cycles, which occurred even when a government was succeeded by another of the same coalition of parties. Thus Portugal's business cycles were often more pronounced than those of its main trading partners especially the EC, as Table 9.1 shows. The first government led by the social democrats stabilized inflation in 1980, and the second government of the same coalition expanded during the world recession of 1981-82. A coalition government of socialists and social democrats stabilized during the world boom in 1983-85, a stabilization that involved a belated increase in domestic interest rates. We see from Figure 9.1 that relative income per capita fell in 1983. Thus the second liberalization-stabilization package, unlike the first, reversed the catching-up process. This time the consequences of the frozen public sector were too strong to be offset by a fall in real wages. The effect was exacerbated because the Portuguese economy was out of phase with the EC economies, a feature that will continue to be a problem as long as fiscal policy is determined by the need to finance too large a public sector borrowing requirement.

That the public sector remained frozen for so long was indicative of the ambiguity of public response towards European integration. Politically it was important to balance European integration against constitutional socialism. Throughout the eight years of negotiations with the EC, conflict between the commitment to EC membership and the commitment to socialism was associated with personal rivalries between the Prime Minister and the President of the Republic. Yet the ambiguity of integration policy in the years after the revolution stemmed from the constitutional system of government as much as from personalities. Because the executive system favored in the constitution was semi-presidential or bipolar, rivalry between the two poles of the executive was hard to avoid. The objective of EC membership was repeatedly emphasized by the socialist leader, Mario Soares, who was Prime Minister from 1976 to 1978 and again in 1983-85, whereas the objective of a socialist economy was defended by the military leaders from the African wars gathered in the Revolutionary Council chaired by General Eanes, the President of the Republic from 1976 to 1985. Furthermore, until the first constitutional amendment abolished the Revolutionary Council in 1982, the loyalty of the President was divided between the electorate who had voted him into office and the army, represented by the Revolutionary Council. Before 1982 the rivalry between the two poles of the executive was particularly crippling. Despite the efforts of a social democratic government in 1980, the economic consequences were progressive expansion of the public sector without any attempt at reform and continuous squeezing of the private sector. Mounting public sector deficits were an inevitable corollary. By the time the Revolutionary Council was abolished and the powers of the President of the Republic were reduced, the problems had become too great to be easily resolved. A succession of coalition governments proved incapable of bringing order to public finances. The process of debt accumulation was becoming explosive, and the rate of inflation reached over 20%.

An anti-inflationary program, coupled with the first steps towards fiscal adjustment, was adopted late in 1985, with the election of another reformist government of the social democratic party led by Cavaco Silva - the minister of finance in the 1980 endeavor. Even though the government's candidate for the Presidency lost, the rivalry between the two poles of the executive subsided shortly afterwards, when Mario Soares became President in February 1986. The popularity of the reformist strategy was strikingly confirmed by the landslide victory of the government in the elections of July 1987.

The government made known during the electoral campaign its intention to revise the constitution and unfreeze the public sector. Its victory set the stage for the second constitutional amendment. This took two years to achieve, however, because to reach the two-thirds majority required passing an amendment, some kind of arrangement had to be reached with the socialists. Perhaps anticipating the difficulty of such agreement, the government decided in November 1987 to sell 49% of the capital of state-owned enterprises to private investors without waiting for the amendment to the constitution. In attempting to divorce ownership from control, the decision made the privatization process more complex but not necessarily faster. The very gradual nature of the privatization process can be viewed as another reflection of the ambiguity of integration policy.

All political parties except the Communists now acknowledge that generalized state intervention was a failure. Yet liberalization has not been accepted easily by any major political force, and the social democrats are anxious to avoid being branded as pro-business. Their preference for some form of social pact was evident in efforts to accommodate the objectives of the non-communist trade union - the General Workers' Union (UGT) - in the disinflation process initiated in 1985. The employers' confederations -especially the Confederação da Industria Portuguesa - also agreed to base wage increases on expected inflation in 1986, but the agreement broke down in 1988 when the inflation outturn was almost twice as high as the target.

Since EC accession the catching-up process has resumed. In 1989 Portugal was forecast to reach a level of GDP per capita slightly higher than Greece. Measured at current market prices and purchasing power standards, Portugal scored 54.4 and Greece 54.1 as a share of the EC average. This was heralded as showing that the country was no longer 'at the tail of Europe'. Yet 'to win 1992' requires structural adjustment as well as faster growth, and both have been hindered by the frozen public sector. The commitment of the government to the single market objective serves as a reminder of the electoral promise to reform the public sector.

The results of the anti-inflationary program initiated with accession should be seen in this light. The government managed to cut the rate of increase of the GDP deflator from 23% in 1983-85 to 15% in 1986-88, while replacing stagnation by a 4% growth rate over the same period. The forecasts of the European Commission reported in Table 9.1 indicate that the performance in 1989-90 will be even brighter, with growth 1 percentage point faster and inflation 3 percentage points slower.

The improvement in inflation was matched by a reduction of the public sector deficit from 11% to 7% of GDP, as shown in the first panel of Table 9.2. The improvement is less remarkable, however, if net unrequited official transfers (mostly from the EC) are excluded from foreign savings, so that capital transfers from the public to the private sector are included in private investment. This approach (which seems to be favored by the IMF) is followed in the second panel of Table 9.2. Conversely, the third panel shows a drastic improvement if state-owned enterprises are included in the accounts of the public sector. This reflects the stringent control that has been exerted on the deficits of these companies. By all three criteria there remains substantial room for improvement. Awareness is growing that the sizeable public deficit coupled with the high public debt, rekindled inflation in 1988, requiring a credit squeeze in early 1989. Unless public finances are reformed, a resumption of stop-go macroeconomic policies will be unavoidable³.

3 Wage flexibility and labor mobility

The labor market has undergone substantial change in recent decades. During 1965-73 more than 100,000 people emigrated to the then six member states of the European Community, but with the change in the immigration policies of the destination countries in 1973-74 Portuguese international migration came to a virtual stop. Moreover, there was massive return migration from the former colonies, amounting to some one million people, equivalent to about one-tenth of the population. At the same time, agricultural employment began a steep decline, which amounted to 30% during 1974-88, though at 21% of total employment it still represents a higher share than in the rest of the Community. The decline in agricultural employment has been compensated by increased employment in manufacturing (21% over the same period) and even more so in services (58%), while employment in public administration grew by 42% during 1981-88. These sectoral shifts have been associated with substantial migration from rural to urban areas.

The activity rate in Portugal (about 44% in 1986) is slightly higher than in the EC overall. The proportion of women in the civilian labor force grew slowly through 1983-88 to reach over 43%, a higher level than the EC average. Paid, registered part-time jobs, however, accounted for only 4% of employees in Portugal in 1986, compared with 12% in the Community as a whole. The proportion of self-employed workers is very high in Portugal, standing at more than 30% in 1988. The figure for average hours worked each week is also higher than the EC average (40 hours against 37.6 in 1986).

Employment has risen in Portugal in every year except 1978, 1982, 1984 and 1985, and unemployment has not been a problem since the late 1950s⁴. The rate did jump up in 1974 and peaked at 8.5% in 1985, then declined to 5.6% in 1988, about half the average for the Community. For new entrants into the labor force Portugal's performance has been closer to that of the rest of the EC, with youth unemployment of 19% during 1984-88 compared with 23% for the EC as a whole. Furthermore, 'underemployment' may be widespread because of the highly bureaucratic public (and private) administration and because of the high cost of adjusting manpower. If the organization of the economy changes, there will be a large potential for workers to be released to undertake new activities. These workers, however, are generally poorly trained and lack special qualifications. Comparatively low wages may attract investment in new industries, which may increase the demand for labor.

During most of the post-revolutionary period real wages have been quite flexible. In the wake of the revolution trade unions and other workers' organizations, which had been docile under the corporatist regime, became very strong and engineered steep real wage increases. Rivalries between the communist and the non-communist union (UGT), however, undermined their bargaining power. At the same time, because they had become accustomed to a stable exchange rate, workers were unaware of the erosion in real wages that currency depreciation was now causing. This allowed the stop-go policies of the various governments to result in falling real wages during most of the post-revolutionary period.

To assess the amount of real wage adjustment and the contribution of devaluation, measures of wage gaps are sometimes used. Such measures provide only a rough approximation, but given the size of the shocks that the Portuguese economy has encountered they can shed useful light. An index of the actual labor share with base 1973 = 100 is presented in Figure 9.3, together with one measure of the 'warranted' share based on the observed consumption real wage, an elasticity of substitution between labor and capital of one half and the small country assumption, that is to say infinite trade elasticities. The Figure shows the gap closing during the two stabilization episodes. The lack of response of nominal wages to exchange rate devaluation was essential to the process of closing the gap, whereas the initial success of the 1985-87 anti-inflationary program induced wage moderation afterwards⁵.

Making internal and external balance a function of the wage gap and the output gap, roughly measured as the deviations of output from trend, we can divide the space of the two gaps into four 'zones of economic unhappiness,' which feature different combinations of internal and external imbalances. In Figure 9.4 the locus of internal balance coincides with the vertical axis and the locus of external balance is downward-sloping: from a point on it a greater output gap (arising, say, from excess demand) requires a smaller wage gap (arising, say, from wage moderation). For one measure of the wage and output gaps, Figure 9.4 plots the observed combinations, so as to elucidate the nature of the adjustment process. The measure of the output gap is given by the deviations from a 2.8% trend growth over the 1973-88 period. The wage gap reported in Figure 9.4 does not make assumptions about trade elasticities. It simply computes the labor share allowing for changes in the terms of trade assuming that the share of non-traded goods in the consumer price index is one-fourth. The base year of 1973 is taken as a position of internal and external balance. The combination of an external deficit and unemployment in 1974-75 is followed by deficit and accelerating inflation in 1976-77 and surplus with inflation after the stabilization of 197&-79. The renewed expansion of 1980-82 turns the surplus into deficit until the stabilization of 1983-85. The expansion of 1986-88, initially characterized by a surplus and disinflation, is now turning towards inflation and deficit⁶. The plots in Figure 9.4 do not exactly follow the observed sequence but demonstrate wide variations in the wages and output gaps, suggesting an adjustment process where real wages bore the brunt of adjustment as the public sector continued to expand.

A degree of flexibility has been maintained in the labor market despite the 1976 law, which confined lay-offs to the most extreme cases. Employers have outflanked the law by hiring employees on renewable short-term contracts (usually 6 months). These employees accounted for an average of 12% of total employment in the period 1983-88. Short-term contracts enable employers to adjust to changes in economic performance, and their existence means that the labor market is not so much rigid as segmented. The announcement in 1987 of the intention to replace the 1976 law by an allegedly more flexible one did not prevent short-term contracts from increasing in 1988.

The impact of the completion of the internal market on the labor market will depend on how long wage levels in Portugal remain significantly below those elsewhere in the Community and on how fully freedom of movement of labor is applied. Portuguese emigration shows a very high elasticity to foreign per capita income. Since the differential between incomes in Portugal and in the developed EC countries is still large, completion of the single market is likely to be followed by emigration. 'The pattern of emigration in the long run depends on how workers perceive the evolution of wages in the home country, since expected higher future home wages lower the emigration rate. If workers believe that home wages will not approach the levels of other EC countries for some time, they will emigrate and perhaps return after wages have adjusted.

With total freedom of labor to move where it likes, new (domestic and foreign) investment will not go to industries that can survive only on the basis of cheap labor, because investors will expect that by the time the investments mature the cost of labor will have risen relative to alternative locations for these industries. If the EC partner countries however maintain their present immigration policies beyond the single market deadline, investment in industries that gain from the use of cheap unskilled labor is to be expected. The period for recouping the investment will be shorter the faster wages converge on EC levels.

Pressure in the labor market for qualified workers is going to be significant even if there is no free mobility, because the level of qualification of the labor force is very low compared with the other countries of the EC and the educational system has been slow to upgrade. Immigration at the top levels of skills is thus likely.

4 Trade liberalization and competitiveness

The world recession induced by the second oil shock in 1979, the strong dollar and the Third World debt crisis did not interrupt Portugal's catching up with the EC until 1983. Between 1980 and 1982 Portugal grew at an annual rate of 3% while the EC stagnated. Yet the fall in the terms of trade was 3% a year between 1980 and 1984, and it was compounded by the rise in world interest rates. The belated adjustment to these two adverse shocks was carried out in a second stabilization package with the IMF, in 1983-84. Then, between 1985 and 1988, Portugal's merchandise terms of trade rose by 22% or 5.8% a year. This improvement, which was due to the falling dollar as well as to the decline in oil prices, was shared by other EC countries, but, because of Portugal's great openness, it had a stronger effect. It allowed the government to pursue an expansionary policy without sacrificing its anti-inflation stance, at least in 1986 and 1987.

This fast-growth policy fuelled an investment boom, which turned out to be much greater than the forecasts in the PCEDED. The public sector behaved as planned so the current account was worse than forecast in 1987. Yet the external consequences remained under control because of the good behavior of exports and sizeable inflows of capital. The favorable trend in competitiveness since 1976 is evident from Figure 9.5. In terms of relative unit labor costs (labeled RULC) it was reversed in 1981 and has deteriorated slightly since accession. The indicator based on unit costs (named RUC), which allows for changes in the relative user cost of capital, shows a decline from the 1983 peak. This shows that the policy of low real interest rates that was pursued because of the high public debt burden did not manage to lower the user cost of capital relative to the main trading partners. Indeed, capital costs are likely to be higher for private firms, which do not have favored access to credit from nationalized banks. Since accession, the major explanation for the good export performance has been not exchange-rate policy but the response of traditional export sectors to a stable macroeconomic environment.

This was also the case before the revolution, when the escudo was pegged to the dollar. Even though it appreciated in nominal terms relative to the pound and the franc and inflation picked up after 1967, exports to EC and EFTA markets remained competitive. Indeed, the relative neglect of the colonial markets in Africa was already visible in the 1960s on both the export and the import side. Since then the EC and, in particular, neighboring Spain have become even more dominant export and import markets (Table 9.3). As the colonial issue became more overwhelming politically, it was losing economic importance. While free trade with the African colonies rose in political and declined in economic significance, European integration became of more economic but less political importance.

The divergence between political objectives and economic imperatives helped to perpetuate the ambiguous response to external liberalization. It was exacerbated after the revolution. State-owned enterprises remained long after their external competitiveness had been eliminated in favor of private exporters. This dual structure reappears in different guises throughout the period. Thus manufacturing is dominated by textiles -particularly cotton - and oils refining, each a good example of the two distinct segments of manufactures. The first is labor-intensive and oriented to exports and the domestic market; the second is more capital-intensive and import substituting. These characteristics are evident in the sample of industries (identified by their NACE Code as well as by name) reported in Table 9.4⁷.

The first segment, which began as an export enclave in the 1960s, includes cotton (NACE Code 432), knitwear (436), clothing (453) and footwear (451), metal products, small tools and equipment (316), pulp and paper (471), glass (247), ceramics (248) and cork products (466). Textile exports represent 25% to 30% of total exports. Except for pulp, economies of scale internal to the firm are small. The average value of production per firm is less than or equal to the UK or Italy but larger than Spain. In the textile industry, however, Portugal and Spain have a dualistic structure with a large number of very small family-run firms and a few large, more competitive and more export-oriented firms.

The firms nationalized after the revolution belonged to the second segment. Import-substituting, state-owned firms are either monopolies or had significant market power in the domestic market in 1983. For these industries the average size of firm, in terms of output and employment per firm, is greater than or equal to the average size of firms in the UK, Italy or Spain. Such is the case for oil refining (140) and cement (242). The iron and steel industry presents smaller firm size. Before EC entry these industries were protected from international competition through several devices, including quantitative restrictions such as those on oil products. They were also favored in terms of government procurement. Foreign investment was restricted or even forbidden. As a result industries in the second segment were heavily concentrated.

In spite of this dual structure, simulations which apply different versions of the single market program to the eight industries listed in the top panel of Table 9.4 indicate that social welfare in Portugal would tend to rise because consumers' gains would more than offset producers' losses. To interpret the results reported in the left-hand panel of Table 9.5, note that the experiments involve 'tariff equivalent' reductions, which reduce direct trade costs, and increase intra-EC trade flows. If the number of firms is fixed (short-run), an increase in imports will increase competition and lower prices, so consumers will gain. If the fall in profits offsets this, welfare falls. In the long run, profits are restored to base values through the exit of firms from the industry. Firm scale is increased and average costs decline, but a greater concentration brings higher prices so that consumer surplus falls. This experiment underestimates what is involved in the transition to EC membership combined with the completion of the single market, because the major effect of 1992 is to replace segmented national markets by an integrated one. In that case, referred to in the right panel of Table 9.5, firms must charge the same producer price, although consumer prices might still diverge because of differences in trade costs. The same difference between short-run and long run welfare effects is present here. Since firms will reallocate sales towards the domestic market (where they previously charged a higher price), production rises in net-importing countries. But since the (common) price-cost margin is reduced, prices are lower and therefore welfare effects are larger if firms cannot price discriminate. As a percentage of base consumption, welfare gains in Portugal are larger than in Spain, which in turn tend to be close to those for the entire EC⁸.

The positive results in Table 9.5 hide substantial losses in output and profits. More detailed case studies of textiles and clothing, the success story of EFTA phase, provide grounds for cautious optimism. The first industry accounted for 25% of manufacturing employment in 1985. Even though it is only a negligible export, it produces a major input for clothing, which in turn accounted for 22% of manufacturing employment and 17% of exports in 1985. This export share rose from 13% in 1981 to 22% in 1987. The predominance of small-scale firms in these two industries is clear from the total number, respectively 684 and 1,467 in 1985. However, their evolution has been different. In the input sector, concentration has been increasing together with investment in machinery. Labor costs have become less overwhelming and labor productivity in 1985 was about 40% higher than in 1977. Price trends suggest an upgrading from low- to high-grade fabrics. In the output sector, concentration has been falling, investment has been modest and labor productivity increased by only 20% between 1976 and 1985. Even though manual work predominates, the low cost of labor enabled textiles and clothing to achieve a very substantial improvement in their shares of the EC market between 1980 and 1985, respectively from 1 to 2% and from 2 to 4%.

Yet, although Portugal already exported (and re-exported) 102% of output in 1985, the improvement in market shares for clothing were even more impressive for Turkey which may have replaced Portugal as a 'pajama republic'. Moreover, evidence of quality upgrading for Portugal comes from comparison of unit prices which are twice as large as Turkey's but only about two-thirds of Italy's. Wages per man hour in 1986, on the other hand, were 2 US dollars, about the same as in Hong Kong, half those of Spain and one-fifth those of Italy. If the transition to higher grades is fast enough, this trend will not be threatened by competition from outside the EC⁹.

Is the fact that comparative advantage has been strongly at work, despite the adverse fiscal environment, another example of market forces overcoming the consequences of ambiguous public response to external liberalization? Up to a point. The optimism must be qualified when a more aggregative view is taken, grouping sectors according to the growth of demand (measured by apparent consumption) in the EC, the US and Japan. So-called 'strong-demand' sectors such as office and data processing machines, electrical and electronic goods, and chemicals and pharmaceuticals have been growing at over 5% p.a. since the 1960s. Growth of about 3% p.a. identifies 'moderate-demand' sectors, such as rubber and plastics, transport equipment, food, beverages and tobacco, paper and printing products and industrial and agricultural machinery. Finally, so-called 'weak-demand' sectors, such as metal products, miscellaneous manufactures, ferrous and non-ferrous ores, textiles, leather and clothing, and construction materials have experienced a growth rate of less than 2% p.a. in recent decades. These groupings have been thought robust enough to explain the strength of Germany's export performance and to predict the costs of adjusting to 1992. To the extent that interindustry trade is associated with weak-demand sectors, the single market will imply falling profits in labor-abundant countries.

Portugal's industrial structure according to these criteria is presented in Tables 9.6-9.9. Higher and growing output shares for weak-demand sectors, especially in the very dynamic trade with Spain and the other EC countries, the declining index of intra-industry specialization in strong-demand sectors, and rising employment and investment in weak-demand sectors, all seem to indicate that significant adjustment costs are to be expected from the necessary restructuring of Portuguese industry. If firms do not invest in differentiated products, export upgrading will be too slow and the industrial structure will remain fragile.

The same message emerges from the recent acceleration of foreign direct investment in Portugal. The accumulated rate of increase from the 1980-85 average to 1986-88 was 153% (79% for manufacturing) and the share of the EC partners rose from 48% to 67% during the same period. Although foreign direct investment is likely to be a major force in Portugal's industrial restructuring, the level is still below the Community average. Table 9.10 shows that foreign direct investment is mostly directed to weak-demand sectors. Table 9.11 reveals the changing composition in terms of sectors and type of operations: the attractiveness of services (financial or otherwise, including real estate) and of existing firms is quite pronounced.

Trade and investment opportunities are thus being exploited, but fundamental policy choices are being postponed, perhaps until the external environment - and in particular the common external tariff- are not quite as favorable to Portuguese exports to other EC markets. The pattern of trade competitiveness shows features of both inter-industry and intra-industry specialization, so that the economy's response to the 1992 challenge remains as ambiguous as public policy¹⁰. Eventually a better domestic environment will be needed to contend with the effects of continued worldwide trade liberalization.

5 Banking regulation and competition

A better domestic environment requires a strong financial market. This implies not only less interference on the part of the government but also a clearer notion of what the public sector is. Moreover, the major obstacle to a more transparent fiscal policy - revenue seeking - will be eroded by the increased factor mobility that the single market will entail¹¹.

In highly competitive and unregulated markets, banks tend to avoid the financing of medium- and long-term investment and to concentrate instead on short-term operations. Where control of corporate behavior through mergers or take-overs is restricted (and banks are more regulated), firms tend to be committed to their group's commercial bank, which in turn adopts a more flexible approach and a longer-term view. In Portugal the creation of industrial groups which were able to provide their own finance was a consequence of the excessive financial regulation of the 1960s. Tight regulation of credit - with ceilings established on an individual bank basis - made it essential for an emerging industrial and financial group to avail itself of a commercial bank. Most of the seven 'family' groups were indeed called by the name of their respective commercial bank. Because of the close links of these groups with the government, the competitive fringe of new conglomerates did not manage to bring about industrial and financial restructuring.

Had this fringe been successful, thriving firms might have been able to shop around for more attractive sources of funding, but the financing of the export enclave of small manufacturing firms did not operate in this competitive way. Most firms outside groups, deprived also of the option of borrowing abroad, had to finance their long-term investment through their own resources or through revolving short-term loans. The oligopoly situation of the seven groups, together with the comfortable external position, explains why non-monetary financial intermediaries failed to develop. Excessive regulation was not limited to commercial banks¹².

Because of their diversified earnings, large industrial and financial groups, such as those identified with the 'seven families', do not require a financial market for investment. In the 1960s the managers of the commercial banks at the core of these conglomerates were nurtured in a type of financial intermediation where most of the operations were internal to the group. As the groups could do without a financial market, after the nationalizations of 1975 they probably took the same view of the nationalized sector as a whole. The closing down of the stock market, shortly after the revolution, reinforced this perception. At the same time, the abnormally high levels of gold and foreign exchange reserves changed its nature. Urged by politicians to put banks 'at the service of the people', the managers saw those reserves as collateral against which the nationalized enterprises were borrowing. It can be shown that when there is imperfect monitoring of projects by banks, a rise in collateral makes borrowers more likely to default. To compensate for this decline in the expected return of a loan, the bank selects riskier borrowers and projects, which cause expected bank returns to fall. To avoid a fall in profitability, the bank will prefer to ration credit, even if there are no macroeconomic disturbances.

Credit rationing in equilibrium is exacerbated in a disequilibrium situation characterized by binding credit ceilings on private firms. In an inflationary environment, where every borrower, especially the government, faces negative real interest rates, the disequilibrium effect may be very strong. Financial repression made private firms more dependent on bank credit at a time when retained earnings were low and there was no substitute in the stock or bond markets. The heavy dependence of firms on bank credit made the financial system more fragile because some of the debts were not expected to be repaid to the banks. This reinforced the lack of incentives for creditors to monitor borrowers. It is thus no surprise that after the stock market was revived in 1987 it attracted mostly firms with insufficient retained earnings and with low collateral.

During the expansion of 1981-82, with high wage inflation and controlled prices, profits and retained earnings fell. When interest rates were raised in 1982, the adverse selection effect towards riskier borrowers (who are less reluctant to pay higher rates) was probably offset by a less binding constraint on credit ceilings. This allowed banks a better mix of borrowers and projects. But, in so far as deposit rates were administratively fixed, higher lending rates increased the intermediation margin and reduced the profitability of nationalized banks. Arrears and bad debts accumulated, peaking at 15% of credit to non-financial enterprises and individuals in 1986. At the time non-performing loans were three times as large as the equity of commercial banks. The system is effectively bankrupt: bad debts fell to about 11% of credit in early 1989, but the figure in Spain is 4%. The figures would be much higher for the nationalized banks, which are also saddled with the need to provide for the pensions of their numerous staff¹³.

Paradoxically, the role of nationalized banks in collecting hidden taxes may have helped stabilize the system. The hidden taxes were collected through their excessively wide intermediation margins and passed on through their forced purchases of public debt. Since nationalized banks were acting as collectors of implicit taxes, depositors could be confident that the state would bail them out¹⁴.

Ten years after the great nationalizations the government finally authorized new entrants, both domestic and foreign, into the banking sector, even though some of them were direct competitors of the nationalized commercial banks. The combined share of the eight nationalized commercial banks in total commercial bank credit fell from 98% in 1979 to 94% in 1985, on account of the rise in the share of foreign banks from 2% to 6%. The share of the eight fell again to 91% in 1987 because of the rise in the share of private domestic banks from zero to 3%. In terms of total assets the new commercial banks were then equivalent to an increase in the total number by one more bank of equal size: in other words, the number equivalent Herfindahl index rose from 8 to 9 between 1985 and 1987.

The new banks avoided holding public debt instruments other than Treasury bills, which had only been introduced in 1984. They were also reluctant to lend to the troubled state-owned enterprises. This meant that they were less exposed to bad debts and suggested a contrast between 'clean' and 'tainted' banks that might prompt depositors' runs on the allegedly riskier banks. The share of private commercial banks is still small, especially on the liabilities side, but their spectacular growth has had a strong demonstration effect on banking competition, stronger than the mere increase in the number of players.

Instead of attacking the debt overhang of nationalized banks, however, the government decided to impose further regulation on the new banks through a ceiling on deposit rates and a (retroactive!) increase in equity requirements. Because the criterion for establishing credit ceilings was related to a bank's capital, there were significant incentives to overcomply, so that the initial negative reaction to the retroactive capital requirements subsided. The only long-term solution is purification of the 'tainted' banks rather than contagion of the 'clean' ones. On the eve of privatization, however, the government announced a series of mergers within the nationalized banking sector, whereby 'clean' (mostly investment) banks were saddled with 'tainted' commercial banks. This procedure, comparable to the engineered mergers of the nationalization period, met with resistance from managers of the well-run nationalized banks. As a result the government has withdrawn the proposal and has said that the state will buy the bad debts of nationalized banks. To the extent that depositors believed that ultimately the state would bail out the nationalized banks, this was indeed the appropriate measure. The consequent increase in public debt will in part be offset by the proceeds from the privatization of most of the state-owned enterprises that were nationalized in 1975, now that the Constitution has been amended to allow it.

Once again, in the light of the single market objective, the process of catching up needs to be coupled with economic restructuring. The banking sector, through privatization or other means, will have to absorb the overhang of inefficiency. This is unlikely to be completed before 1992 though a faster readjustment cannot be ruled out. As with trade and industry, the pressure for financial readjustment is coming more and more from outside - and it is visible in the booming Spanish financial market. No excessive regulation in Portugal is likely to last without severe damage to financial development, because business will go across the border to Spain. Against the slow evolution of Portuguese nationalized banks, financial restructuring in Spain began several years ago and continues, in the form of mergers among large banks, often encouraged by the monetary authorities.

6 Fiscal adjustment in a frozen public sector

Credit ceilings have been the major instruments of monetary control. In their present form they date from 1978, the time of the first stabilization package. Not surprisingly the allocation of credit to the private sector has been squeezed because of the debt behavior of the enlarged public sector. Since the demand for credit exceeds what can be afforded by a responsible monetary policy, the potential excess liquidity ends up in the banking system. The rational response on the part of the commercial banks would be to turn down deposits that cannot find their way into loans. To pre-empt this response the Central Bank has created interbank securities markets, where through repurchase agreements, banks invest in securities that are part of the Central Bank's portfolio. The rate of remuneration of the banks' excess liquidity has been determined through an auction system, which makes it a kind of 'market' interest rate. Since liquidity purchases have been increasing rapidly - in part because of capital inflows from abroad - new instruments have been created and a more active interest rate policy has been pursued in the interbank market. Yet, in 1988, the net effective interest rate on the public debt held by the Central Bank turned negative.

Restrictive measures were introduced in March 1989 which relied mostly on tighter credit ceilings (including much higher reserve requirements) and avoided the large increases in interest rates that would be required to cool off demand.¹⁵ The real interest rate on domestic public debt rose from -9% in 1980-82 to -3% in 1986-88, but the difference with the rate of growth remains negative and large, as shown in Table 9.12, column 5. Although a further rise in interest rates would increase the burden of public debt, and lower the profits of the Central Bank, it might also reduce the scope for implicit taxation, as argued below.

In July 1989 the Portuguese government approved a revised version of the Program to Correct External Imbalances and Unemployment (PCEDED): reduction in consumption and transfers, increases in taxes, and privatization of public enterprises, including banks, should generate primary surpluses sufficient to stabilize the total debt to output ratio before 1992. This revision had been expected since late 1988 and the objective is not very ambitious. To judge from the 1990 budget, however, whether the required fiscal restraint can be enforced is open to question. One prerequisite for any fiscal adjustment strategy, which is lacking is reliable public finance data. The planned size of the yearly adjustment must be credible, but in the absence of data there is no way to monitor the progress of the adjustment effort. Moreover, because some debt is held involuntarily the interest burden calculated by dividing interest payments by the nominal value of the debt is overstated. This is another reason why the PCEDED should be more ambitious. A consolidation at the market rate would imply a substantially lower value of public debt.

Available figures on Portuguese public finances show debt standing at over three-quarters of GDP in 1988, which is already quite large by EC standards. The debt to GDP ratio rises to 100% when 'guarantees' provided by the general government for loans to public enterprises are taken into account (the debt of the electricity company alone was about 20% of GDP in 1987). To the extent that these loans will never be repaid, the existence of these 'guarantees' justifies the notion of a frozen public sector that includes the general government and also state-owned enterprises, including both financial and non-financial enterprises. The so-called 'enlarged' or 'administrative and entrepreneurial' public sector definition was used in the third panel of Table 9.12. Unfortunately its borrowing requirement is difficult to reconcile with debt figures and the reported government deficit.

The accounting framework used here is based on the debt to income ratio rather than on the debt itself. In broad terms this ratio should as a rule be stable in a steady state. In high public debt countries it should probably decline. If interest rates on debt exceed growth rates, the public debt ratio would rise without limit. In such a situation bondholders, fearing that the authorities will try to wipe out the value of the debt through inflation or repudiation, would precipitate a financial collapse. Using such a basis for sustainability calculations implies that bondholders' fears are so strong that they induce the government to announce and enforce a multiannual fiscal adjustment strategy (MAFAS). This is simply a plan to stabilize the accumulation of public debt through increases in revenue and decreases in expenditure such that, excluding interest payments, there is a budget surplus¹⁶. This rule of thumb provides a useful benchmark, as long as the underlying assumptions are kept in mind.

Calculations based on the debt to income ratio indicate an implied, or actual, deficit that has often been much larger than the reported deficit, suggesting the presence of a sizeable hidden deficit. Table 9.12 decomposes public debt as a share of gross domestic product since 1970. The broader measure of public debt used shows an unsustainable situation, with a ratio increasing at 4% a year for over ten years, despite high growth and negative real interest rates. The difference between the implied deficit - the change in the debt to income ratio net of the interest and growth factor - and the reported deficit is in part due to deficient data, but largely reflects disguised fiscal policy.

Unreported lending operations by the Treasury and debt take-over operations by the government (to the benefit of autonomous funds as well as of public enterprises) are acknowledged sources of discrepancy. They make the decomposition of stock accumulation into well-defined flows difficult to interpret on a year-to-year basis but facilitate future fiscal management. Before the revolution the reported surplus was hiding a deficit, except in 1972 when the reported deficit was hiding a surplus. Between 1977 and 1985 - leaving out 1980 because of the debt write-off operation - the implied deficit was on average double the reported deficit, though there were substantial year-to-year variations. This is the same as saying that on top of the reported deficit there was a hidden deficit of equal size. In 1986-88, however, while the reported primary budget deficit was about zero, there was a hidden deficit of 6%.

In Tables 9.13 and 9.14, to interpret better this trend in the total debt to income ratio, public debt is decomposed into domestic and foreign and the domestic into privately and publicly held debt. Total debt accumulation accounts for foreign borrowing by the state and distinguishes the voluntary from the forced holding of public debt. The former is restricted to Treasury bills; all other holdings of public debt count as seigniorage, which is an implicit tax. The reason for using this definition rather than the monetary base is precisely the focus on revenue seeking by the state. Seigniorage revenue from the foreign counterpart of the monetary base accrues to the foreign state. The interest and growth factor is correspondingly restricted to the debt voluntarily held¹⁷.

In Table 9.13 the revenue from seigniorage is given by the net accumulation of non-privately held debt (the tax rate) times the non-privately held debt to income ratio (the tax base). Although the revenue has been falling, the pattern is more erratic than the averages in Table 9.13 suggest.

In 1981 and in 1984 revenue from the central bank was 7% and 8% respectively, and in 1981 and 1985 revenue from the nationalized banking system, including the central bank, reached 9% and 10% respectively. The table shows that the tax base and rate differ substantially depending on what is treated as being in the public or the private sector. The different base and rate for the seigniorage tax reflect the distribution between debt privately and publicly held, since in both cases net foreign borrowing remains the same. If nationalized banks are aggregated with the private sector, the tax base remains at about 20% throughout the past decade, but the tax rate falls from 10% in 1980-92 to 1% in 1986-88. Conversely, if only Treasury bills are taken to be privately held, the rate still falls, from 12% to 3%, but the base actually rises from 34% to 41%. Here only the domestic counterpart of the monetary base is included in the definition of seigniorage. When the change in the monetary base is used, there is a much smaller decline in tax revenue from seigniorage. The difference is related to the strong external position in 1986-87¹⁸.

The existence of implicit taxes and hidden deficits is central to the pattern of macroeconomic adjustment observed in Portugal since the revolution. The adjustment process has largely spared government expenditure. The absence of restrictions on public spending implies larger increases in revenue and resorting to hidden forms of taxation. The accounting framework introduced above was able to uncover disguised fiscal policy. In the same way, taxation analysis cannot neglect the economic incidence of all forms of taxes. To understand fiscal adjustment correctly necessitates going beyond overt incidence¹⁹.

This approach shows the taxation of financial intermediation to be far different from what is prescribed in the tax code. There is an implicit intermediation tax imposed on borrowers and depositors in the banking system. The rate of implicit intermediation tax depends on the spread between the loan rate and the deposit rate net of an assumed intermediation margin, since both borrowers and lenders suffer from excessively large interest margins. The implicit intermediation tax base depends on the slopes of demand for and supply of credit and deposits. These are in turn related to the alternatives offered to residents. In general seigniorage and other concealed taxes will be collected not only from borrowers and depositors but also from bank shareholders. If the binding constraint is the existence of credit ceilings of private borrowers, the tax base is the corresponding stock of loans. Here this is assumed to be the case, even though private depositors may occasionally have been constrained.

To compute the implicit intermediation tax rate, an average rate on deposits can be calculated from the figures for the stock of total deposits and the interest bill paid by banks. The loan rate for the 91- to 180-day maturity (which was administered until September 1988) can be used as the representative rate for loans extended by commercial banks during the sample period. Credit ceilings imply that these loans are a relatively small share of the total assets of banks. Estimates of the excess burden are shown in Table 9.15 for the case where a reasonable intermediation margin is assumed to be 3%. If the tax base is private credit, then the hypothetical implicit intermediation tax 'revenue' is sizeable. It peaks at almost 10% of GDP in 1982, falling to close to 4% in 1987.

The pressure of the single market deadline was not strong enough to affect the design of the 1988 Tax Act²⁰. The most important effect of the introduction of comprehensive income taxation in 1989 may be an increase in the credibility of a future tax reform in which concealed taxation is reduced to a level consistent with external financial liberalization. Then the government, through a multi-annual fiscal adjustment strategy, will be able to commit itself to restoring control over public finances. This commitment is all the more necessary since the doubling of Community structural funds cannot be expected to continue forever after. Moreover, such a large increase is bound to bring additional pressure on public investment expenditure, because of the requirement that recipient countries match these funds to an equal amount.

7 Monetary and exchange-rate policy

Disguised fiscal policy ends up determining monetary and exchange-rate policy as well. But the effectiveness of capital controls needs also to be taken into account. International capital mobility and free trade in financial services, by greatly increasing the competition among banks, is bound to make Portuguese banks unwilling and unable to finance the deficits of the public sector at rates substantially below comparable borrowers. The conflict between banking competition and constitutional socialism affects public sector reform and the process of catching up in various ways. The most immediate is probably the increased cost of collecting the implicit intermediation tax from depositors, borrowers and shareholders. The new behavior began in 1985 with the new banks and is likely to become stronger with the banks privatized in 1989. At the same time, even if the liberalization of capital movements on a Europe-wide scale is delayed, it is difficult to believe that barriers against the movement of capital between Portugal and neighboring Spain will be effective. As barriers to trade in goods are being dismantled in accordance with the transition agreement with the EC, capital controls are becoming less effective.

Most EC member countries have retained some restrictions on international capital movements and so have been placed in a group characterized as having a semi-open capital market²¹. Greece and Portugal, like several developing countries, were in the group with closed capital markets. Exchange controls kept interest rates in Portugal artificially low, indicating that as barriers to capital outflows they operated very stringently. Table 9.16 shows in columns 5 and 6 that the average covered interest differential against the dollar between 1984 and 1988 was 0.6% in Spain and about - 3.0% in Portugal. If more recent observations are introduced and the early period ignored, the difference between the two neighboring countries is maintained but Portugal emerges as much less closed.

In view of the opening up of domestic capital markets required by the single market objective, the public finance situation in Portugal is threatening the sustainability of external liberalization. This is also the case in other EC countries with a high public debt, such as Greece and Italy. Temporarily to maintain a closed domestic capital market may be justified as a transitory protection of inflation tax revenues. Hence it has been suggested that high-inflation EC countries might together pursue financial repression in a 'soft currency club'. These soft currencies would crawl relative to the EMS, so as to stabilize relative prices throughout the EC. It would be difficult to organize a special arrangement of this sort, however, and the case for such a halfway house is not convincing. Tables 9.16 and 9.17 point to some of the limitations of this approach by suggesting that a passive crawl like the one followed by Portugal since 1977 not only may have ceased to alter relative prices, but has also introduced a significant bias in the system.

The crawl bias shown in the first column of Table 9.16 is obtained by subtracting the nominal interest differential relative to the basket peg using the rate of crawl from the realized interest differential. A positive bias implies that the domestic currency turned out to be less attractive than expected. The size of the bias is related to the variance-covariance matrix of exchange rate changes relative to the numeraire, and it falls from 1986 largely because of the greater fixity of exchange rates after the January 1987 realignment within the EMS.

Like Table 9.17, Figure 9.6 suggests that the crawling peg is no longer altering relative prices. The rate of crawl (icrawl) is compared with the change in an effective exchange rate (NEF) in which roughly equal weight is given to the dollar, the pound, the DM and the French franc. It can be seen that the change in cost competitiveness (relative unit costs including cost of capital, as in Figure 9.5 above) is mostly due to the discrete devaluations of the 1978-79 and 1983-85 periods.

The effect on the Portuguese stock markets of the Wall Street crash of October 1987 was certainly stronger and more lasting than would have been expected in a fairly closed market. This observation confirms the importance of financial development as the output shares of insurance, banking and the marketing of securities grow strongly in EC countries, including the less developed. Table 9.16 shows that the capital market has become much more open since EC entry, with the average negative covered interest parity relative to the dollar falling from close to 6.8% in 1985 to zero in 1987 and 1988. The same pattern can be observed in the case of real interest rates, where a positive differential of almost 9% in 1985 becomes a negative differential of 1.5% in 1988, as well as in the decrease in the expected change in the real exchange rate, related to imperfect integration of goods markets. These developments are in line with the substantial capital inflows observed during and after EC entry and the successful anti-inflationary strategy of 1985-86.

Credit ceilings are a feature of closed capital markets, and in Portugal they have proved quite effective during periods of stabilization. In spite of recent rises, interest rates tend to be kept low so as to alleviate the burden of public debt. This shows again how the political element creeps into financial discipline. The only credible measure to end the direct financing of the Treasury by the banks in EC countries with a high public debt may be an agreement among the central bank, the ministry of finance and the spending ministries on a plan of deficit reduction involving both expenditure and revenue, and perhaps including tax reform. This is the essence of a multi-annual fiscal adjustment strategy (MAFAS).

Because of exchange controls, monetary policy will remain much more passive than in the current members of the EMS or in the UK. True, the government's diminished scope for collecting the implicit intermediation tax led the Treasury to announce that it would cease to have automatic access to the central bank with effect from 1989²2. Even if this 'divorce' lasts in the 1990s, it is not clear how far the incomes policy program, which tends to be agreed between the government and the mostly socialist trade union, the UGT, will circumscribe the independence of the central bank.

For the Portuguese central bank to follow the policies of the German central bank would be an important institutional change. How well the operation of monetary policy adapts in preparation for this change greatly affects the assessment of the costs and benefits of joining the EMS. The divorce between the bank and the Treasury needs to be consolidated and monetary policy needs to learn to operate without the strictures of implicit taxes. The announcement in 1989 of the intention to introduce indirect monetary control needs to be implemented soon if it is to be believed. The decision to subject credit to the public sector to the same ceilings as credit to private firms, with effect from I July 1989, is a welcome change in the direction of eliminating the privileges of public-sector borrowing. But it is no substitute for a (credible) MAFAS.

With respect to exchange rate arrangements, the experience of the United Kingdom outside the exchange-rate mechanism of the EMS seems to suggest the advantage for Portugal of exploring forms of association which will enhance the credibility of domestic macroeconomic management without excessive reliance on the policies of the BundesBank. To experiment with some form of wider band, as in Italy and Spain, seems unavoidable. Indeed, the pressure on the United Kingdom brought about by Spain's joining the EMS and the gathering momentum towards economic and monetary union make earlier membership of the system more likely.

The effect of joining the EMS is simulated in Table 9.18 by fixing the escudo to the DM and comparing the real effective exchange rate obtained, with and without price adjustment. Except in 1988, the real appreciation that has occurred since 1985 would have been exacerbated by pegging to the DM. This is because the escudo depreciation relative to the DM was less than the dollar, franc and pound depreciation in 1986 and 1987. Had prices remained fixed, the real appreciation would have been even stronger, reaching 44% in 1988 relative to 1985. Moreover, it would have continued in 1988. These two experiments provide upper and lower bounds for the effect of exchange-rate policy during the 1985-86 disinflation process. Without a passive crawling peg, the real appreciation would have been excessive, yet inflation was partly fuelled by depreciation. These very crude measures suggest therefore that disinflation cannot last without prior control over the level of expenditure, which is impossible without fiscal adjustment. Joining the EMS with largely open capital markets will not reverse inflationary expectations and thus will not be credible unless reductions in the public sector borrowing requirement are expected²³.

For a time in 1987 the pound sterling followed the EMS informally, but demand expansion made this a short-lived strategy. Were Portugal to make a similar attempt it would require a greater degree of monetary and fiscal restriction than in the PCEDED. A shadow exchange-rate arrangement would be less rigid than membership of the EMS, which would tie the central bank's hands. If the MAFAS is so gradual that the public sector remains essentially frozen, shadowing the EMS may be the only feasible alternative consistent with opening the capital markets before 1996. In short, for the immediate future experimentation with an informal peg may be more credible than either keeping the crawling peg or joining the EMS with a wide band, though as time passes shadowing will become a less satisfactory solution. Should the MAFAS not materialize beyond the existing PCEDED, such extra credibility would quickly disappear. The argument for shadowing has been made more pressing by Spain's decision to join the EMS. The great acceleration of trade between the two neighbors is exerting pressure for their financial markets to be integrated as well. This makes it harder for Portugal to pursue an independent monetary policy, even before the beginning of stage one of the Economic and Monetary Union.

8 Conclusions

Portugal's experience makes clear that the expectation of external liberalization does not by itself secure a change in the domestic regime. This was true for labor, trade and industry as well as for finance. Ensuring the durability of the change in policy regime hinges essentially on the credibility attached to the multi-annual fiscal adjustment strategy. Since the existing program for debt stabilization is not explicit about the means to achieve the objective, its credibility may be questioned. Immediate steps should include improving knowledge of the debt situation of the enlarged public sector and, perhaps, rethinking exchange-rate policy, in the direction of shadowing the EMS.

The order in which trade protection and financial protection are removed does matter. It would be preferable for domestic real liberalization to precede external financial liberalization²⁴.

This suggests that the removal of financial protection requires domestic financial liberalization, namely, the recognition of the implicit taxes and administrative controls that are designed to favor borrowing by the Treasury. The ability to bring about an irreversible monetary reform without external pressure may be limited in countries with a high public debt. By requiring the implementation of a multi-annual fiscal adjustment strategy designed to control durably public expenditure, external financial liberalization may be a way of neutralizing the efforts of those groups close to the public sector who lobby for deficit financing and for implicit taxation because these require less adjustment on their part.

The financial regime in Portugal has changed a great deal since private commercial banks began to operate in 1985. Nevertheless, domestic seigniorage and other implicit taxes remain significant burdens on the economy, let alone sizeable sources of government revenue. The pressure of 1992 has not been sufficient to affect the design of the comprehensive income tax introduced in January 1989, but it is raising awareness of the need for a credible multi-annual fiscal adjustment strategy that will go beyond the recently revised Program to Correct External Imbalances and Unemployment (PCEDED). The agreement of the EC to double structural funds by 1992 is having the same effect, both because it cannot be expected to continue forever and because it requires additional public investment.

When the disinflationary process was reversed in 1988, inflation was still substantially higher than the Community average. At the same time, the financial sector is facing a serious overhang of bad debts. Under these conditions, shadowing the EMS seems to be the appropriate signal that the objective of economic and monetary union will not fail in Portugal because of the burden of a frozen public sector. The burden is evident in commercial banks that are providers of grants disguised as loans paid for by the collection of taxes disguised as margins. In Portugal fiscal adjustment is a prerequisite for financial development. More than that, it is a prerequisite for sustained economic growth and structural change. The longer public sector reform is delayed, the faster it must proceed to meet the single market deadline. Even if financial liberalization is as remote as 1996, its impact will combine with the efforts to move towards economic and monetary union and is likely to be felt much earlier. The tradition of ambiguous public response to external liberalization is not an attractive feature of Portugal. The government should abandon it by embedding the 1990 budget in a multi-annual fiscal adjustment strategy, which will meet the 1992 challenges and bring about the desired economic and social cohesion of the European Community.

NOTES

This country study for the CEPR project on 'Economic Integration in the Enlarged European Community' draws on nine background papers identified in the references. Financial support from the Institute of Portuguese Foreign Trade (ICEP) and criticism from project participants are gratefully acknowledged. Joan Pearce was responsible for substantially reducing ambiguity in the analysis and in the text. Special thanks go also to Bill Branson, for his comment, and to Odete da Cunha, who typed the manuscript countless times. The views expressed are personal. They should not be interpreted as those of the Directorate General for Economic and Financial Affairs of the Commission of the European Communities (DGII), where the author is Director of National Economies.

1 The channels of structural interdependence are analyzed in Macedo (1981) and the EFTA phase is discussed in Macedo, Corado and Porto (1989). See also Macedo (1989a), whose Annex contains a discussion of the recurrence of liberalization-cum-stabilization packages between 1977 and 1985, drawing on Corado and Macedo (1989).

2 Government bonds were promised a few years later, but this compensation was regarded by former shareholders as grossly inadequate. The issue became entangled with the legal nature of nationalization and has remained unresolved. The constitutional obstacles to modernization are addressed in Macedo (1984).

3 Freitas (1989) contains further evidence on financial balances and the PCEDED objectives and realizations in 1987-90. Thus annual growth in private consumption was 4.9% rather than 2.9%, public consumption 2.6% rather than 1% and investment 13.7% rather than 7.9%. The 3% difference between actual and forecast domestic demand growth was made up by net imports, so output growth was only 0.5% higher than forecast. Inflation however was almost twice as high 10.5% rather than 5.6%.

4 This was particularly remarkable in 1974-75 when output fell by more than in the EC but employment rose, thanks to a 20% increase in public-sector employment. The model of Barosa and Pereira (1989) tracks well emigration, activity and employment from 1959 to 1985, with a structural break in 1974. Caveats about the data abound.

5 The magnitudes are very approximate because of the unreliability of the data (especially employment data before 1983) and because of the 'wage gap' methodology. The alternatives are explained in Freitas and Macedo (1990).

6 This is the modification of the well-known Swan diagram proposed in Krugman and Macedo (1979). The interaction between the phases of the macroeconomic adjustment process and the gains from trade and factor mobility are taken up in Macedo (1984).

7 According to Leao,Ramada and Reis (1989), the export concentration in EC markets leaves Portugal vulnerable to a loss of competitiveness relative to the newly industrialized countries of the Far East or the emerging market economies of Eastern Europe. While both product differentiation and market diversification are desirable, the pursuit of comparative advantage is difficult to criticize - except insofar as it has been based on subsidies such as the ones included in the PEDIP. If the upgrading called for in Macedo (1984) continues the concentration is not likely to be a serious problem. But the evidence of Leao etal. (1989) is very scanty.

8 Corado (1989) compares the two Iberian countries with each other as well as with Italy and the UK. The Spanish country study emphasizes the 'EEC cum 1992 shock'. Corado and Leite (1989) extend the analysis in Corado (1989) and consider the case where Portugal does not remove protection. The welfare gains are somewhat larger than in the segmented market case but smaller than in an integrated market. The sectors covered are only NACE 316, 345, 350 and 471.

9 See note 7 above. Neven (1990) identifies Portugal and Greece as the EC members relying more heavily on comparative advantage based on labor-intensive goods. Unlike Greece, however, Portugal displays comparative advantage in natural resource-based goods - such as paper pulp (NACE 471). This highly concentrated industry is the third case study in Leao et al. (1989). There are 3-4 firms but over 70% of output is exported and investment per worker has increased tenfold and labor productivity has doubled since 1977. The predominant export is eucalyptus, which has sometimes been seen as ecologically unsound. (See D. Smith 'An eucalyptus success story'. Financial Times, 11 October. 1989.) Similar evidence of export dynamism can be gleaned from other industries, especially telephonic equipment, cereals and biscuits, as well as railway equipment. Buigues and Ilzkovitz (1988) propose a methodology, which is applied by Goncalves (1989) to these and other 'sensitive' industries.

10 This goes back to our discussion of the effects of inter- and intra-industry trade, in the context of Portugal's accession to the EC. The notion in Macedo (1984) was that differentiated products and foreign direct investment from outside Europe could replace emigration. Krugman and Venables in this volume show how a small country's comparative advantage in labor-intensive manufacturing may be overwhelmed by market access and size at both high and low levels of trade barriers. They also claim that labor mobility reinforces centripetal forces and may thus hurt the periphery. See Mexia (1989) for a statement of Portugal's industrial and trade strategy after accession: differentiated products and diversified markets.

11 The importance of the revenue-seeking motive was already noted in Macedo, Corado and Porto (1989) in connection with the introduction of the 1985 sales tax-which was supposed to compensate for the loss in tariff revenues. See also note 20 below.

12 If only commercial banks were too tightly regulated, there would be incentives for other financial institutions to create money. Note that when commercial banks are not regulated enough, they become prone to panics and disasters. Corbett (1989) draws this lesson with Eastern Europe in mind.

13 Branson (1986) suggests that the outcome could have been either a collapse or a stable solution. He cites interviews giving much higher figures than the ones from Banco de Portugal, BoletimTrimestral, March 1986 reported in the text. Many bad debts are not listed as non-performing loans, and there is no disaggregated data available from the banks' reports. During the stabilization episode of 1977-79 bad debts were kept outside the credit ceilings, so that there was a perverse incentive running against the adverse selection effect. These points are elaborated upon in his contribution to this volume and in his comment on this paper.

14 The proposed implementation in Portugal of a scheme such as the Federal Deposit Insurance Corporation in the US could only exacerbate the bias towards bad loans, even on the part of heavily exposed banks. There is already an excess of insurance against bank failures, based on the gold reserves of the Banco de Portugal as well as on an implicit guarantee by the taxpayer.

15 According to Silva (1968), prime minister since 1985, the bias was already present in 1966. Freitas (1989) discusses the growing ineffectiveness of credit ceilings and the increase in reserve requirements in March 1989 (from an average rate of 6% to a single rate of 17%), together with the recent measures to eliminate the privileges of the public sector as a borrower.

16 The original version of the PCEDED approved in 1987 did not include a MAFAS. The current version compares the base case with a 'no adjustment' scenario. The first yearly Plan for Public Debt (PDP) was also approved in 1989.

17 The revenue perspective followed here suggests this definition, but at the cost of preventing comparabilitywith Table 9.12. If income growth were spread over the entire debt (including the debt involuntarily held), the interest and growth factor would remain the same. See Gaspar and Macedo (1989). Tables 9.12 and 9.14 are not directly comparable but it is clear that excluding foreign borrowing and debt not held in the nationalized banks makes a substantial difference. Also the average implied deficit is the same for both criteria, and the interest factor is negligible for Treasury bills. Since the reported deficit has a mostly political content, it is interesting to note that it is only slightly larger on average than foreign borrowing.

18 In his comment on this paper, Branson provides the following reader's guide to Tables 9.12-15: "The expression for the growth rate of the ratio of debt to income (db) is given in equation (3) of my paper, repeated here for reference:

(1) db=(r-n)b+d-s.

Here r is the real borrowing rate, n is the real growth rate, b is the ratio of debt to GNP,d is the ratio of the non-interest deficit to GNP, and s is the ratio of seigniorage to GNP. Equation (1) can be rewritten as an expression for the implicit deficit, as

(2) d=db-(r-n)b+s.

In Table 9.12, the implied deficit in column (6) is calculated as the change in the debt-income ratio, our db, in column (1) less the interest-growth factor, our (r —n)b, in column (5). This is the implicit deficit net of seigniorage, (d - s). For example, in 1980-82, the implicit deficit in Table 9.12 is 7 percent of GNP, compared with a reported deficit of I percent. To obtain the total implicit deficit, inclusive of seigniorage finance, we add central bank seigniorage tax revenue/ Y from Table 9.13,2 percent in 1980-82, for a total 9 percent. The financing of this deficit can be obtained from column (5) of Table 9.14, which gives the IIT component. For example, in 1980-82, the foreign component was 3 percent of GNP, and the IIT was 6 percent.'

19 Gomulka (1989) identifies a 'hidden interest rate subsidy’, which rose to 25% of GDP during the hyperinflation of June/December 1989. Hillman (1989) dwells on 'soft budget constraints' in Hungary, a point emphasized in the Greek country study and by Rodrik in this volume.

20 The modernization of the tax system began with the introduction of a value added tax in 1986, was followed by the introduction of a comprehensive income tax on I January 1989 and is to be completed by the reform of capital taxation. Without the abatement of concealed taxation fiscal adjustment would be incomplete on the revenue side. The importance of simplicity is evident in the proposal for tax reform of Macedo and Gaspar (1989), which involves a single marginal tax rate.

21 The interpretation uses the approach of Jeffrey Frankel as discussed in Macedo (1990). A complete set of data is in Macedo and Torres (1989).

22 The accounting framework used in Section 6 was originally applied to Italy, where the situation was similar before the so-called 'divorce' between the central bank and the treasury in 1981. See Macedo and Sebastiao (1989) and Beleza and Macedo (1988) on the Portuguese experience. Klein and Neumann (1989) discuss the German and UK experience of central bank seigniorage.

23 The real appreciation resulting from joining the EMS is dealt with by Branson and by Krugman in this volume. See also Gaspar's comment and Gaspar and Macedo (1989).

24 The argument is made in connection with Greece, Spain and Portugal in Macedo (1990), where further references are cited.

REFERENCES

Background papers 

Barosa,J.P. and P.T.Pereira (1989). 'Labor Supply in an Environment with Emigration: An Empirical Study of the Portuguese Case', August. 

Corado,C. (1989). 'Portuguese Industry and the Effects of EC Membership', August. 

Freitas,M.L. (1989). 'Portugal PeranteoMercadoUnico: As ReformasMacroeconómicas no Processo de Ajustamento', September.

Freitas, M.L. and J.B. Macedo (1990). 'The Portuguese Wage Gap Revisited', January.

Gama,A.C. (1989). 'VantagensComparativaspara Portugal', August. 

Leao, E., P. Ramada and A.B.Reis (1989). 'EvoluçãoRecenteePerspectivas Futuras do Sector ExportadorPortuguês de Tecidos,Vestu+ario e Pasta de Papel', August.

Macedo, J.B. (1989a). 'The Political Economy of Portugal's European Integration 1949/88'.August.

(1989b). 'Financial Integration and Disguised Fiscal Policy in Portugal', August. 

Mexia, A. (1989). 'Conclusões', September.

Other references

Beleza, L.M. and J.B. Macedo (1988). Implicit Taxes and Credit Ceilings: The Treasury and the Banks in Portugal', Working Paper No. 106, Nova University of Lisbon, December.

Branson, W. (1986). 'Portugal's Entry into the European Communities: Challenges and Opportunities', draft paper of a World Bank Report. 

Buigues, P. and F.llzkovitz (1988). 'The Sectoral Impact of the Internal Market',

draft paper. Commission of the European Communities, DGII, July.

Corado, C. and A.N.Leite (1989). Industrial Location and Foreign Investment in the Iberian Peninsula', draft paper presented at the Second European Research Workshop on International Trade, Bergen, June.

Corado, C. and J.B. Macedo (1989), 'Competitiveness under Liberalization cum Stabilization Packages (LCSP): the Experience of Portugal 1977-1985', Nova Economiaem Portugal (Barbosa Festschrift), Nova University of Lisbon, pp. 541-58.

Corbett,J. (1989). 'Policy Issues in the Design of Banking and Financial Systems for Industrial Finance', draft paper presented at DGII, Commission of the European Communities, Seminar on Poland and Hungary. 

Gaspar, V. and J.B. Macedo (1989). 'The Case for Budgetary Restraint in Portugal: Short-Run and Long-Run Arguments', draft paper written at Conference on 'Economic Integration in an Enlarged European Community', October, Delphi.

Gomulka, S. (1989). 'Monetary Policies in Poland 1989-90', draft paper presented at DGII, Commission of the European Communities, Seminar on Poland and Hungary.

Goncalves, P. (1989). 'LesEnjeuxSectorielsduMarché Interieur pour 1'Industrie Portugaise', draft paper for the Commission of the European Communities, DGII-V project on the Internal Market, Associação Industrial Portuguesa, October.

Hillman, A. (1989). 'Macroeconomic Policy in Hungary and its Microeconomic Implications', draft paper presented atDGII, Commission of the European Communities, Seminar on Poland and Hungary.

Klein,M. and M. Neumann (1989). 'Seigniorage: What Is It and Who Gets It?', Institut fur InternazionaleWirtschaftspolitik,Universitat Bonn, Discussion Paper No. B-124.

Krugman, P. and J.B.Macedo (1979). 'The Economic Consequences of the April 25th, 1974 Revolution',Economia 3, 455-83, reprinted as Economic Growth Center Paper No. 299, Yale University, 1981.

Macedo, J.B. (1981). 'Portugal and Europe: The Channels of Structural Interdependence', in J.B. Macedo and S.Serfaty(eds), Portugal since the Revolution: Economic and Political Perspectives. Boulder, Colorado: Westview Press.

(1984). 'Portugal and Europe: Dilemmas of Integration'. In T.Bruneauetal. (eds.) Portugal in Development: Emigration, Industrialization, the European Community, University of Ottawa Press.

(1985). IntegraçãoEnropeia:Fim do PrincipioouPrincipio do Fim? Economia 9 ,5-19.

(1990). 'Financial Liberalization and Exchange Rate Policy in the Newly Integrating Countries of the European Community', forthcoming in Bergamo Conference on 10th Anniversary of EMS, edited by P. Ferri: Macmillan.

Macedo, J.B.,C.Corado and M.L.Porto (1989). 'The Timing and Sequencing of Trade Liberalization Episodes in Portugal', Working Paper No. 109, Nova University of Lisbon, January.

Macedo, J.B. and V. Gaspar (1989). 'UmaReforma Fiscal para 1992: Imposto Linear UnicosobreoRendimento', Nova Economia em Portugal (Barbosa Festschrift), Nova University of Lisbon, pp. 381-402.

Macedo, J.B. and M. Sebastiao (1989). Implicit Taxes and Public Debt: the Portuguese Experience', European Economic Review 33, 573-79.

Macedo, J.B. and F. Torres (1989). Interest Differentials, Financial Integration and EMS Shadowing: Portugal and a Comparison to Spain', draft paper presented at Conference on Portugal and the Internal Market of the EEC, Lisbon, October.

Neven,D. (1990). 'EEC integration towards 1992: some distributional aspects.' Economic Policy No. 10, 5, 13-62.

Silva,A.C. (1968). 0 MercadoFinanceiroPortugues em 1966, Lisbon: Gulbenkian Institute of Science.