THE ECONOMIC INTEGRATION IN CENTRAL AMERICA: DOOMED TO FAIL?

Five Central American countries, Costa Rica, El Salvador, Guatemala Honduras, and Nicaragua, formed the Central American Common Market (CACM) in 1960. For about two decades the countries performed very well in terms of standard macroeconomic indicators. On paper, although not always in reality, the CACM was a very far-reaching organization and certainly the most sophisticated example of economic integration ever in the Third World. It has also widely been regarded as one of the most successful ones. Tensions between and within the participating countries grew up almost from the start, however, and starting in the late sixties the common market gradually begun to disintegrate when Honduras withdrew from the organization (Brock 1989, p. 309). Increasing political unrest in the region was obviously one reason for this, although far from the only one. In fact, the disturbances might have been a symptom, rather than the cause, of the economic problems ofthe CACM. We argue that many ofthe difficulties that emerged were a direct or an indirect consequence of the policies pursued. The final blow to the common market came with the "debt crisis" in the early 1980s. The extensive intra-regional trade that had developed within the CACM collapsed, sometimes even being replaced by outright trade embargoes (Weeks 1989, p. 24). Today the CACM is, for all practical purposes, a paper organization, even if some of the institutional framework created remains. Nevertheless, the importance of reviving the CACM has regularly been emphasized (cf. Irvin and Holland 1989, p. 1, Guerra-Borges 1989, p. 59, Central America Report, May 19, 1989), although another strand of thought, typically represented by the most important external financiers (IMF, World

Bank, Inter-American Development Bank, and US AID) tends to downplay the integration, emphasizing a global division of labour instead (cf. Central America Report, April 28, 1989). It is widely recognized, however, that a revival would not be an easy task; the economic factors contributing to the collapse of the CACM remain. Furthermore, political tensions in the region have seriously aggravated the problem.
The purpose of this paper is to give an interpretation of the macroeconomic developments during the period ofthe CACM, that is from 1960 on. To do this we first have to give some historical, institutional, and social background. The aim is to monitor the effects of the integration and to give an interpretation of what was going on under the surface of the encouraging macroeconomic data. Our thesis is that the ultimate failure of the integration to hold together was more or less an inevitable result of the economic, political, and institutional mechanisms at work in the Central American countries. Hence, although there are significant external factors behind the economic collapse (cf. Lizano 1989, pp. 290-291), the internal ones appear to be decisive.
We will, as far as possible, treat the region as a whole, not emphasizing differences between the five republics unless this is important forthe argument.

The Organization and Policy of the CACM
The attempts at integration in Central America should be seen against a background of history which has fostered a feeling of community between the five countries (cf. Blomstrom & Lundahl 1989,p. 5, Castillo 1966. The region was administered as a single entity by Spain, and even formed a federation shortly after independence, in the 1820s and 1830s. The federation was short-lived, however, and in spite ofthe fact that the Central American countries have much in common, the political relations between them have, from time to time, been strained, sometimes to the point of war. Under colonial rule, the republics had been developed as self-contained entities. The economic and social structure ofthe five countries show striking similarities despite great differences in size, income level and population density (cf. Barraclough & Scott 1987,p. 39, Fraseret al. 1987, p. 146, Castillo 1966,p. 6, Delgado 1978. Repeated attempts at establishing some kind of Central American unity have been made through the years in spite of the difficulties. All these efforts were more idealistic than practical, however, and could not envisage practical means for achieving their objectives. Hence, none ofthem lasted for more than a few years (Castillo 1966, pp. 67-68). Nevertheless, they probably laid a psychological foundation for the integration process of the 1960s and 1970s. Concrete ideas of closer economic cooperation between the Central American countries started to turn up once more after World War II, when it began to become clear that the traditional export-oriented strategy, based on primary products, although quite successful for a long time, could not take the countries much further (cf. Castillo 1966, p. 66).
The philosophical background for the organization consisted primarily of the ideas of the United Nations Economic Commission for Latin America (ECLA) -subsequently organized within the so-called structuralist paradigm (Hunt 1989, ch. 5) -propagated from the early 1950s on by Raul Prebisch (1950 and1959) and others. According to this paradigm it was necessary to break loose from the dependence on the industrialized world. Import substituting industrialization and integration between less developed countries (LDCs) were considered effective means of development (cf. Hunt 1989, pp. 141-143). Nevertheless, it should be noted thatthe ultimate goal forthe CACM was never introverted development. On the contrary, the idea was that greater interaction with the world economy could be achieved in the longer run, when the new industries had matured (cf. Bulmer-Thomas 1989, p. 69).
At the outset, the prospects for classical import substituting industrialization did not look too good forthe single countries: the technological level was, typically, low, capital was scarce, and the governments were weak and reluctant to mobilize capital via taxation. On the demand side the markets were small in each country, partly because ofthe unequal income distribution 1 • Furthermore, the risks of negative welfare effects within the CACM were substantial, as the conditions for successful integration were not met at the time of the establishment of the union. In 1960, intra-regional trade accounted for only 6 per cent of total trade in the Central American countries and the industrial base before establishment of the union was insignificant. In addition, the five economies werepredominantiycomplementary, rather than competitive, with each other, with the exception of primary products (Bulmer-Thomas 1982, p. 238).
The emerging common market had quite a long gestation period. In the 1950s industrial output grew more than the economy in general. In spite ofthat, the expansion was not sufficient to absorb the growth in urban populations. Meanwhile, the small and fragmented domestic markets did not provide enough incentive to prospective industrialists (Castillo 1966, p. 59). All this contributed to the insight that a regional ratherthan a national view was called for. The fundamental goals for the integration venture were, on the one hand, to accelerate industrialization by increasing the size ofthe "domestic" market and, on the other hand, to facilitate full participation in the world economy (Lizano 1989, p. 280). It was probably of some importance that a good deal of institutional strengthening took place in the early postwar period, such as the establishment of central banks, and the emergence of social legislation and a tariff policy (Castillo 1966, p. 69).
Cooperation between the five countries proceeded cautiously and gradually. Instead of the grand visions of earlier attempts at unification the integration process proceeded more on a step-by-step basis, where solutions to limited practical problems were sought.
A political organization, the Organization of Central American States (ODECA) was founded in 1951 a Trade Subcommittee was established the same year, and a multilateral treaty establishing a customs union was signed in 1958 (Bulmer-Thomas 1977). Earlier in the 1950s, bilateral free trade agreements were signed between individual countries (Cline & Delgado 1978, p. 1). The General Treaty of Central American Economic Integration was signed in 1960, (effective from June 4, 1961(Guerra-Borges 1989), formally establishing the CACM; Costa Rica signed the agreement later, in 1962 (Bulmer-Thomas 1987b, p. 76).
The target period for trade liberalization was five years (Guerra-Borges 1989, p. 48)2. Part of the agreement was the foundation of several important regional institutions, among them SIECA (Permanent Secretariat of the General Treaty of Economic Integration) for co-ordination of the integration process, CABEI (Central American Bank for Economic Integration) for financing of infrastructure projects, and CCC (Central American Clearing House) for the settlement of intra-regional trade obligations. Moreover, an Economic Council, with the senior ministers of economy as members, was established (Guerra-Borges 1989, p. 54).
In 1970, FOCEM (The Central American Monetary Stabilization Fund) was created to facilitate coping with temporary balance of payments problems. The five members of the CACM and, additionally, Venezuela and the U.s. official development aid organization, USAlD contributed to the fund (Brock 1989, p. 309). Due to the failure of FOCEM to cover the very big imbalances from the late 1970s on, a new fund, the Central American Common Market Fund, was created in 1981. In reality the institutional framework remained weak (cf. Fuentes 1989, p. 90), presumably due to the high priority of extraregional trade as compared to intra-regional trade.
The framework for the Central American integration was not formed without difficulty. In particular, concern was generally felt about what was going to happen with government revenues, possible negative impacts on already existing facilities, and the risk of concentration of industrial activities to one or two countries (Castillo 1966, p. 81). With the benefit of hindsight it can now be said that the concern, especially as to the third point, was not entirely unjustified.
The main policy instruments forthe CACM countries were three: the use of free trade zones, where industries were allowed to import intermediate products and capital goods duty-free, a common external tariff (CET) , and different fiscal incentives. The target for many ofthe policy measures seems to have been foreign enterprises. Today about one third ofthe industrial sector in Central America is foreign-owned (Guerra-Borges 1989, p. 60, see also Cline & Rapoport 1978, p. 283). The purpose offoreign direct investment has usually been the exploitation of "protection rents" -excess profits due to trade barriers -(cf. Reynolds 1989) in the domestic markets and not too much in the name of creating an export base (Guerra-Borges 1989, pp. 58,60). The long run aim was to develop an industrial sector strong enough to allow exports of manufactures to the rest of the world (cf. Bulmer-Thomas 1989, p. 67).
The internal tariffs were gradually abolished. At the moment the General Treaty became effective (1961), 81% of all dutiable trade was freed from restrictions; by the early 1970s the figure had risen to 96% (Guerra-Borges 1989, p. 54). F orthe external tariffs the general pattern was that the CET raised the nominal and effective rate of protection on consumer goods while the rates on capital goods were lowered (Bulmer-Thomas 1987b, p. 79). This, as is typical of import-substitution, strongly favoured the production of consumer goods. Fiscal incentives, in effect subsidies, were also used, concentrating on industries processing domestic primary products and capital goods industries (Guerra-Borges 1989, p. 56). The incentive policy developed into more of a problem than an asset, however, as the countries tried to outbid each other to lure new firms to establish operations within their borders (Lizano 1989, p. 286).
Realizing that economies of scale and polarization of production may be a problem on the small internal market, particularly when some important intermediate goods are concerned, a regulated establishment of so-called integration industries was planned (Guerra-Borges 1989, p. 48). The industries were supposed to be set up in all five countries and were to be granted a virtual monopoly position for ten years within the CACM, including duty-free imports of inputs and free access to the regional market, subject to governmental supervision of their conduct (Bulmer-Thomas 1989, pp. 80-81, Castillo 1966. Not much ever came of that plan, however. Only three such industries were founded (in tires, glass, and caustic soda).
In spite of obvious attempts at coordinating their development strategies, a spirit of competition regarding e.g. economic incentives granted to importsubstituting industry remained between the countries, and the question whether the CACM, as an organization, had an industrial policy at all, is debateable (cf. Guerra-Borges 1989,pp. 48, 52, Bulmer-Thomas 1987b,p. 80, Reynolds 1978 3. The Economic Developments in Central America since 1960: An Overview Traditionally, the engine of the economies of Central America has been the primary sector, agriculture comprising between 3 5 and 45 per cent of GNP in 1950. Cash crops, such as coffee and bananas and after the second world war also cotton and sugar, provided the foundation of the economy. (Some other natural resources, such as metals, ores and oil do exist, but not in great abundance (Castillo 1966,pp. 12, 156).) In many of these activities the Central American republics had quite a strong world market position, and this exportoriented strategy granted high rates of economic growth. Still, exploitation of natural resources cannot warrant expansion forever -despite the fact that further diversification took place (cf. Cappi et al. 1978, p. 324) -particularly if the linkage effects with the rest of the economy are small.
Morover, the dependence on one or two export commodities rendered the Central American countries very sensitive to international fluctuations. Along with quite impressive GDP growth rates on the macroeconomic level (4,6% on average per year for the period 1950-1960, see Figure 4), several problems became more and more prevalent. After the Korean War, the barter terms of trade took on a declining trend which was to go on for decades 3 , in fact up till the present. Partly because of this trend, the trade balance has been a continuous headache (Bulmer-Thomas 1977).
Mainly due to the rapid growth of population and lack of alternative employment, significant deterioration of the position of the peasants took place. The income distribution was getting skewer and growing tension emanated from the conflict between the urge to increase production of foodstuffs and the incentive ofthe landowners to expand their supply of cash crops (cf. Bulmer-Thomas 1987b, pp. 2-3). As a rule, cultivation of foodstuffs was crowded out to less fertile land (Castillo 1966, p. 38).
Environmental decay, for instance taking the form of deforestation, and consequent erosion, gradually became visible. The availability of food has apparently been inadequate ever since the tum ofthe century (Castillo 1966, p.42, Barraclough & Scott 1987,p.19)4. Thefactthat,aspointedoutbyWeeks (1989, agricultural production has not been able to keep pace with the population growth in any ofthe five countries points in the same direction. In  Thomas (1982. p. 237) andUnited Nations (1991. p. 142) VI the light of the balance of payments situation, stagnating domestic production is probably an indication of declining nutrition standards as well.
The problem may be more one of absolute poverty than one of short supply offoodstuffs, a considerable part ofthe population being too poor to buy sufficient quantities or an adequate variety offood. The estates were often kept large only to keep the peasants offthe land, and thus secure cheap labour for the plantations (Castillo 1966, p. 40, Fraser 1987. From time to time more or less half-hearted land reforms have been implemented in all five countries, without much success in the long run (cf. Caceres & Irvin 1989, p. 170, Arcia 1989, p. 47, Barraclough & Scott 1987. A consequence has been a substantial migration from the rural areas to the cities (cf. Bulmer-Thomas 1977) Nevertheless, the rural population continued to grow rapidly. Despite that the agricultural (active) population increased by 47 per cent during the same period, the corresponding figure for the "urban" (non-agricultural) population rose by no less than 146 per cent. The total population in the region almost tripled between 1950 and 1990. The rate of increase has regularJybeen around 3 %, showing only a modest sign of declining, as demonstrated inFigure I.This of course implies an immense pressure on land and capital resources, while migration gradually created an explosive situation in the urban areas.
The outward-oriented economy provided, despite the problems just touched upon, a solid basis for diversification of the economic base in the Central American countries (cf. Castillo 1966, p. 66). Traditionally, the economies of Central America had followed the same "model" and had been stable in the sense that the rate of inflation was low and, consequently, the exchange rates could be kept stable and the currencies fully convertible (cf. Bulmer-Thomas 1977 and1987b, pp. 1-4). The latter factor facilitated the workability ofthe Common Market, in effect eliminating the exchange rate risk in intra-regional trade (cf. Brock 1989, pp. 318-319) Industrialization within the CACM was carried out according to the traditional import substitution model, starting from relatively simple consumption goods. Intermediate and capital goods continued to be imported from third countries. This was facilitated by the structure of the tariff protection. The average nominal rates established with the CETwere 106% for nondurable consumer goods, 36% for industrial inputs, and only 11 % for capital goods. Individual incentives in the form of tariff exemptions, however, were quite common (Guerra-Borges 1989, p.55 The average tariff for all products was 48% in 1965, a figure to be compared with its counterpart from before the general treaty: 42% (Castillo 1966, p. Although the emphasis on the production ofthe three types of products did change, particularly in the favour of intermediate goods, between 1960 and 1978, near the de facto end of the CACM, non-durable consumption goods remained the biggest group. Figure 2 gives an idea ofthe development for the region as a whole. (There is some, but not very great, variation between the five countries. For details, see Guerra-Borges (1989, p. 62).) The share of extraregional imports of total domestic demand for the different types of products reveals a similar pattern (Figure 3).
The overall dependence on extra-regional imports remained more or less the same over the whole period up to the late 1970s, while evidently some import substitution in the category of intermediate goods did take place. The ratio of extra-regional imports of manufacturing to apparent consumption was also fairly constant, around one third, between 1970 and 1985 (Willmore 1989, p.52).
The primary product exports continued to guarantee an inflow offoreign exchange and continued to be the backbone of the economy, comprising about 60 per centoftotal exports for all five countries as late as in 1985 (Luiselli 1989, pp. 69-70). (An indication of the significance of the traditional sector is the strong correlation between the savings ratio and the international coffee price, found by Caceres and Irvin (1989, p. 183).) Cash crops are still very dominating in extra-regional exports. Significant diversification has occurred, however, with respect to both products and export countries. According to Irvin (1988) the main export commodity of each country accounted for roughly 70 per cent ofthe region's foreign exchange earnings, while about 80 per cent of its total trade was with the United States. In 1978 both those figures had been halved.
The overall growth rates of GDP were among the highest in Latin America during the the 1950s and 1960s. The 1960s, in particular, stand out as a period with exceptional growth. This extraordinary performance was, however, evidently of a one-time nature. It was due largely to the expansion of mutual trade opportunities and the heavy public investments in infrastructure needed to support the emerging industrial sector (cf. Pira & Fuentes 1987, Blomstrom & Lundahl 1989). Still, it should be noted that the high growth rates cannot be explained by the effects of the CACM only. According to Bulmer-Thomas (1979), only 10 -25 per cent -according to different estimates of the total growth in the sixties was due to integration. The contribution of the CACM to the growth of manufacturing seems to be more evident, though (cf. Willmore 1989).
The rapid growth continued in the 1970s, although the pace was already slowing down. A summary of the average growth ofGDP and manufacturing in different periods is given in Figure 4.
The growth rates of GDP are much less impressive, however, in per capita terms. As a matter off act, the average income level per capita has clearly deteriorated in the 1980s. In 1985, the GDP per capita levels in the five CACM countries corresponded to those already attained in the late 1960s or early 1970s. In fact, even total GDP declined in real terms during much of the early 1980s. The very low rates of investment are perhaps an even greater cause for concern. In the 1980s the rate of growth in gross investment has been negative, sometimes strongly negative in all five countries for several years (Weeks 1989, p. 37). This gives an indication of outright capital destruction going on.
Economic growth is regularly accompanied by structural change. Figure   5 gives a rough idea ofthe structural changes resulting from (and causing) the high growth rates in Central America. As manufacturing grew at a faster pace than GDP throughout the 1960s and 1970s, the ratio of manufacturing output to GDP continued to increase in all CACM countries, reaching a peak of 18 per cent for the region as a whole in 1980.
The growth of intra-regional trade, which was almost entirely a result of the integration, was very rapid in the 1960s and 1970s. The structureofforeign trade also changed, naturally, as a result of industrialization and free intraregional trade: the relative share of imported consumer goods went down, while the share of capital, and in particular, intermediate goods increased. To what extent the emerging manufacturing industries represented sectors with genuine comparative advantages is a crucial question, discussed in detail by Cline and Rapoport (1978).
They also investigated the amount of incentives -in the form of effective tariff protection and tax exemption -different industries received in relation to their comparative advantage. Their results indicate that the "traditional" sectors (e.g. foodstuffs, textiles, and footwear) seem to have reaped the greatest benefits, while in particular metal products, machinery etc. seem to have been discriminated against (Cline & Rapoport 1978, p. 290).
Although the results are not quite clear-cut and probably should be interpreted with a good deal of caution, this may be one reason for the subsequent stagnation in the progress of industrialization, due to the apparent fact that industries where (static) international competitiveness (as measured by comparative advantage) was already great were able to obtain additional support from the system of incentives. The scope for X-inefficiency was likely to be very considerable, whereas potentially competitive industries remained unsupported.  19601970197519801985Sources: For 1950 The local and regional markets were -not too surprisingly -the overw he lmingl y most important destinations of manufacturing output (Bulmer-Thomas 1979, Willmore 1989. This development shows up in dramatically increasing volumes of intra-regional trade? The top was reached in 1980 for the absolute value of trade, but the relative share of intra-regional trade culminated as early as about 1970. The latter development is shown in Figure  6. In 1970,75 per cent of manufacturing output in the CACM was sold in domestic markets, 16 per cent was exported within the intra-regional market and 9 per cent to the extra-regional market. In 1985, these figures were 83%, 9% and 8%, respectively. Much of the new trade was of the intra-industry type. The share of exchange of manufactures for manufactures rose from 38 per cent in 1961 to 62 per cent in 1971, and also at a more dis aggregated level substantial intra-industry trade could be found (Bulmer-Thomas 1979). While not necessarily a negative phenomenon in itself, as growing intra-industry trade could also be a sign of increased competition in the intra-regional market, this certainly suggests that the CACM did not produce the degree of specialization which may have been necessary for producing internationally competitive goods. Such inefficiency was opulently spurred by the high barriers provided by the Common External Tariff and other arrangements.
Looking at the growth rates of exports fort he period 1968-1974, Bulmer-Thomas (1979 found that there was a distinct tendency for exports to the "rest of the world" to grow faster than the intra-CACM exports, in spite of the bias against such exports inherent in the import substitution strategy. This may be taken as an indication of the growth potential in the CACM being exhausted. It is questionable to claim that the original idea of participating "more fully" in the international economy (Lizano 1989, p. 280) was fulfilled, because the overwhelming bulk of the extra-regional trade remained primary products. With the benefit of hindsight we now see that the efforts to diversity their exports geographically did not save the CACM countries from the crash of the early 1980s.
The new manufacturing sector suffered from most of the problems generally attributed to import substitution. The markets were small, partly as a consequence of the small employment effects of an excessively capitalintensive industry, and partly because the income distribution was very uneven (a fact that was enforced by the weak employment effects). The need to import capital and intermediate goods gradually increased the sensitivity to external disturbances due to the fact that the exports of agricultural goods in effect (neglecting foreign borrowing) was decisive for the importing capacity (cf. Pira & Fuentes). FIGURE 6. Ratio of Intra-Regional Trade to Total Trade, 1950-1986 1950 1960 1970 1977 1980 1982 1984 1986 Sources: For 1950-70, calculations of the authors based on Bulmer-Thomas ( , p. 244), for 1977( -86, Brock (1989 While the economies superficially went on developing fairl y satisfactorily, economic tensions were building up, leading to the final collapse ofthe CACM in the early 1980s, along with increasing social unrest and political friction between the countries. In fact, the general view ( cf. Bulmer-Thomas 19 87b, p. 89) is that the dynamism in the CACM had been lost by the time Honduras left the organization (1970). The "easy" phase of import substitution, production of consumption goods, was over.
It took, however, about ten more years until the disaster was plain for everybody to see. Central America was rescued from the economic turmoil of the 1970s by a lucky coincidence in the form of price increases ofthe principal export goods (coffee, bananas, sugar, and meat) occuring in tandem with hikes in the price of oil (cf. Blomstrom & Lundahl 1989, pp. 14-15). With less luck the crisis could have come much earlier. In 1979 the final blows to the integration were delivered: the second oil crisis, rising international interest rates, and the social and political upheaval culminating in the fall ofthe Somoza and Romero regimes in Nicaragua and El Salvador, respectively (Bulmer-Thomas 1987b, p. 139).
The long-run trend in the terms of trade had been unfavourable during the entire existence of the CACM, as Figure 7 clearly demonstrates. For the period between 1978 and 1981, however, they alone fell by 30%, while the cost for servicing the external debt rose simultaneously (Bulmer-Thomas 1987b, p. 143). Since the trade volumes did not change much, an inevitable result was a sharp increase in external indebtedness (cf. figure 9)8. The problems were exacerbated by capital flight, a reaction to increasing political violence, an increasingly overvalued exchange rate, and a general feeling that a crisis was approaching (cf. Pira & Fuentes 1987, FitzGerald & Croes 1989. Deteriorating terms of trade and declining income levels brought about a vicious circle, which spawned severe problems for the integration process. Payment difficulties soon emerged and, probably because extra-regional imports were considered more vital to the economy, intra-regional trade was hit hard, as can be seen from Figure 8. The development of intra-regional trade was also hampered by the accumulated outstanding trade debts (US$ 0.7 bn by july 1986) between the member countries, especially Guatemala and Costa Rica (creditors) and Nicaragua and El Salvador (debtors).
Between 1980 and 1985 the transactions through the Clearing House (CCC) decreased by almost two thirds (for more exact figures, see FitzGerald & Croes 1989, p. 152). The savings ratio fell-probably because of capital flight -which of course did not make the adjustment any easier (cf. Bulmer-Thomas 1987b, p. 142). Partly because ofthis, in vestment activity was rather depressed during the whole 1980s.  Unemployment, although high, did not change dramatically during the 1980s, astonishing enough. Figure 10 shows the development. The available figures probably do not tell the whole truth, however, since disguised unemployment would not be captured by the official figures.
Thus, in spite of obvious progress, something in the development model was evidently wrong, although what that "something" was is subject to some controversy. As Bulmer-Thomas stresses (1987b,p. 2) the policy makers have, by and large, been left with the problem of envisaging a new developing strategy for the region without knowing exactly what was wrong with the old one. The crumbling of the CACM has, in effect, led to a renaissance of the earlier strategy based on the export of primary products (Bulmer-Thomas 1987b, pp. 20-21).

Social and Political Development
The social and political conditions in central America are on the whole a direct consequence of the economic system based on exports of primary goods. The expansion of the cash crop sector put great stress on the rural community almost from the start, and the colonial social structure superimposed on the agricultural sector created a system with very concentrated economic and political power (Castillo 1966, pp. 19,23). Evidently a crucial reason for smallholders not to go into cash crops at the time was that successful entrepreneurs in that field needed complementary resources, such as financing, technical and commercial know-how etc., to which only the wealthier part of the population had access (Castillo 1966, pp. 30-31).
Much of the present social tension stem from the time of the Great Depression in the early 193 Os. At that time, the first phase of economic growth, based on exports of primary goods, came to an end. The system had then become more and more deficient. It did not provide enough employment, nor an adequate level of consumption or sufficient links in the production chain. Moreover, investment necessary to sustain the growth did not come forth (Castillo 1966, p. 53).
Very authoritarian governments were gradually established, after considerable political turmoil, to guarantee the viability of the export-led model. This perpetuated the power of the land oligarchy, however, while the situation for the peasants grew increasingly difficult, as the distribution of income continued to become more uneven (with the possible exception of Costa Rica, which has a more equitable distribution from the start because of a more equal distribution of land). The availability of cheap rural labour  1960 1970 1975 1978 1980 1982 1984 1986 Source: For 1950-60, calculations of the authors based on Bulmer-Thomas (1982, p. 244), for 1970-86, Fuentes (1989 e:J provided an incentive for rapid expansion of the area utilized for cash crops, whereas the cultivation of foodstuffs was neglected. The record of rapid increases in overall income per head thus conceals very dualistic economies and extensive poverty among the rural population. A study on the development of "land pressure", measured as arable land (net of land used for export agriculture) relative to the economically active population indicates considerable deterioration during the period of the CACM. Not unexpectedly, the situation is worst in EI Salvador, but the trend is similar everywhere. The consequence for real income is likely to be most serious for smallholders and tenant farmers. The pressure of migration must be very strong under such circumstances. Bulmer-Thomas concludes, paradoxically, that the success, not the failure, of the export-led model is the source of social and political problems in Central America (Bulmer-Thomas 1987b, pp. 37-39).

1950
A characteristic trait in the history ofCentra1 America after 1960 was the dual system of an industrialization program envisaged by the CACM, and the continuing importance of the traditional primary sector dominated by the old land oligarchy. This coexistence was not without problems (cf. Bulmer-Thomas 1987b, pp. 20-21), and the economic dualism was matched by a political one. Consequently, the degree of repression had to be very high in order to keep the society together ( cf. Bulmer-Thomas 1987b, p. 211), again with the exception of Costa Rica. The suppressed conflicts finally exploded into political violence and rate of urban unemployment civil war in the late 1970s.
Hence, news reports from Central America during the last decade have not in general concentrated on the economic developments, disastrous as they were. Rather, the social and political problems in the region have been the focus of interest. It is true that these features have strong implications for the economy, being partly a reason for the economic problems, but also, most probab ly, a consequence of an economic strategy that proved inconsistent in the long run with peaceful social development.

Why Did the CACM Break Down?
To be able to understand the ultimate failure of the CACM in the early 1980s, we need the support of some trade and integration theory amended with some institutional, political and social considerations. (See Viner (1950) and Lipsey (1957Lipsey ( , 1960 for the basic theory of customs unions. Regarding effects of import substitution, see e.g. Bhagwati (1988), Bruton (1989) or Krueger (1988).) In the light of this theory the CACM was doomed to fail. Source: Feinberg (1989Feinberg ( , p. 271) 1981Feinberg ( 1982Feinberg ( 1983Feinberg ( 1984Feinberg ( 1985Feinberg ( 1986 The exact timing of the stalemate was of course affected by external factors, but these factors could hardly have turned a potential success into a failure. The industrial base in the five countries was very small before 1960. Thus there could not be much trade creation. Instead, purchases from extra-regional sources were substituted for intra-regional imports, viz. trade diversion (trade diverted from more to less efficient producer) or trade suppression (more efficiently produced imported products substituted for domestic products) takes place 9 • As noted earlier, the CACM was an exercise in regional import substitution, with very much of the inherent consequences we are used to encountering in countries relying on that strategy. This does certainly not mean that import substitution as such is an inferior strategy. On the contrary, it can easily be shown in a framework of changing comparative advantage that import substitution is a natural part ofthe structural change (Blomqvist 1991). It is the usual way of pursuing this type of policy, however, with heavily protected domestic markets and a strong bias against exports, which tends to create problems, especially if protection is extended to industries with obvious comparative disadvantages. Particularly important was the excessive capital intensity in industry (due to the tariff structure and to fiscal incentives), which implied rather small employment effects, which in tum tended to preserve the skewed income distribution (cf. e.g. Blomstrom & Lundahl 1989, p. 14)10. Another problem was thatthe saturation level for easily attainable import substitution targets was relatively low, even considering the five economies together. This is partly a consequence of the unequal income distribution, which at least did not become more equitable during the era of the CACM. Much of the private consumption was conspicuous, with small margins to expand (cf. Pira & Fuentes 1987).
Finally, the import substitution strategy changes the structure of imports towards capital and intermediate goods. This makes it much more difficult to cut down imports in cases of balance of payments difficulties. Hence, import substitution tends to exacerbate the dependence on imports instead of reducing it.
A natural next stage for import substituting industrialization would be manufacturing of capital and intermediate goods. This second step in the import substitution strategy is much more demanding, however, as is generally recognized (Colman & Nixson 1986, p. 293). For one thing, the technology required is more advanced. Secondly, the market for such goods is limited even in large countries. In the CACM, the generous incentives offered by the local   Saborio (1990. p. 282) Note: For Guatemala. the 1980 and 1982 figures represent the rale of urban unemployment ~ authorities also favoured external imports, rather than local production, of capital and intermediate goods. In general, but especially in the case of more advanced production, industrialization apparently was hampered by the inability ofthe state to raise sufficient investment funds for infrastructure, education etc. (e.g. Irvin 1988). Instead, much of the available financing was channeled into export agriculture (Irvin 1988, Bulmer-Thomas 1987a. When assessing the relevance of the theoretical arguments above, we may first of all note that the standard theory of integration is of limited value. There seems to be widespread agreement that the trade diversion effects dominated in the case ofthe CACM (Bulmer-Thomas 1987b, p. 86, Willmore 1976. Theoretically, the assessment of the welfare effect of trade diversion is based on allocative distortions only. Iftrade diversion leads to employment of idle or underutilized resources, as was evidently the case in Central America, the impact effect may well be positive, (cf. Bulmer-Thomas 1987b, pp. 77-78, Castillo 1966. Moreover, there may be dynamic effects facilitating an accelerated rate of development (cf. Delgado 1978, p. 21). According to Cline (1978, p. 110-112), the welfare gain were on the orderof3 to 4 per cent of GDP annually (estimated for 1968-72), about half of which was due to dynamic effects. In particular, traditional trade creation and trade diversion effects were relatively unimportant compared to other static and dynamic effects (Cline 1978, p. 111).
The problem with the CACM seems to have been the unequal distribution of the benefits between the participating countries, not a lack of benefits as such. This problem was indeed perceived, however, and measures were taken to alleviate it. Among these measures was the idea of "integration industries" mentioned above, preferential financing through CABEI, and special treatment concerning fiscal incentives to newly established industries (Lizano 1989, p. 285). The Central American Bank for Economic Integration (BCIE) did favour the less industrialized countries, but this was not apparently enough to persuade those countries (particularly Honduras) that they were fairly treated.
The institutional weakness of the CACM can easily be seen in this context. How a redistribution of benefits between CACM countries should have been carried out is somewhat debateable, however, since the inclusion of dynamic effects changes the results obtained considerably. The conventional conclusion, confirmed also by Cline (1978, pp. 94-95), has been that Honduras was the country that gained least from the CACM (which was the main reason for its departure from the CACM in 1970), butthis does not hold unambiguously after including the dynamic effects. On the other hand, the latter effects are no doubt subject to more difficult estimation problems. Thus, all interpretations should be made with caution.  1984 1985 1986 1987 1988 Source: Saborio (1990, p. 283) Note: For Guatemala and El Salvador, the index represents the real average wage, and/or the latter,1976-78=100 $ The most important message of Cline's estimates is perhaps his finding that all five countries did gain something, although in different proportions (Cline 1978, pp. 110-112». The pattern of disintegration after 1970 and the political discussion at the time seem to confirm the general picture of relative gains, as given in e.g. Cline (1978), even ifthere may be room for discussion on details of the calculations and estimates. The smaller the markets were for products manufactured with relative ease in the region, the sooner the limitations for successful import substitution would be reached (cf. Bulmer-Thomas 1987b, p. 78, Castillo 1966. Extra-regional trade seems to have retained its importance during the whole history of the CACM. In fact, the export incomes generated by extraregional exports were positively correlated with intra-regional imports (Bulmer-Thomas 1987b, p. 82). Thus, exports of primary products remained the engine ofthe economies. For industrial production and intra-regional trade to grow, agricultural extra-regional exports had to grow, too. In a sense, intra-regional trade comprised goods of rather low priority, apparently due to the lack of real interdependence (Castillo 1966, p. 100), which could be done without if necessary.
According to Bulmer-Thomas (l987b, p. 82) the CACM has increased instability, since intra-regional trade relied, indirectly, on extra-regional trade. In fact, one of the criticisms of the CACM development model is that the basic economic and social structure remained largely unchanged, as the new industry only added to the existing structures without changing them. The industrial sector, hence, became just an "additional layer" on top of the old structures (cf. Pira & Fuentes 1987, Irvin 1988).
Rather high rates of effective protection may have been crucial for two reasons. On the one hand, fairly high profit rates had to prevail in the manufacturing sector for capital to be diverted there from primary production (even ifthe latter seems to have been starting to exhibit diminishing returns after the war). This is the usual difficulty for resource-rich countries to industrialize (cf. Blomqvist 1991). The "Dutch Disease" effect may have been more important still: a strong traditional export sector maintains a foreign (real) exchange rate too low for new industries to cope with (cf. e.g. Roemer 1985, pp. 234-352).
The willingness of domestic capitalists to engage in manufacturing may therefore vary with the prices of primary commodities. On the demand side, however, the foreign exchange earned from exports of traditional goods were a prerequisite for the demand for manufactured products to come forth. It is noteworthy that the growth rate of the agricultural sector was very weak during the 1980s, very much along the same lines as the development of the whole economy (Weeks 1989, p. 35). While this is most likely a consequence of the severe social unrest in the region during this period, it may also suggest that the limits to expansion of this sector are being reached.
Protection was also needed to lure foreign capital into the region. It is rather obvious that the multinational enterprises that established themselves in Central America did so to exploit "protection rents". Thus, difficulties in the regional market provided a strong reason for both domestic and foreign capital to pull out.
The importance of protection rents can clearly be seen from the fact that the new manufacturing firms did not engage themselves in much extra-regional exports, and also very seldom founded subsidiaries in more than one country (Bulmer-Thomas 1987b, p. 95). Similarly, domestic businessmen, by and large, stuck to small plant sizes and old technology, not even aiming at exporting theirproducts outside the region (Castillo 1966, pp. 100-10 1). While this may partly be due to a lack of experience and training, it is obvious that the structure of tariff protection did not encourage exports (where the effective rate of protection may easily have been negative).
The "shortage of foreign exchange", in the sense that imports of capital and intermediate goods must be on a certain level to keep production going (cf. e.g. Krueger 1988, p. 359), hence tended to become more and more prevalent.
The problem of external balance did not, however, appear serious during the 1970s, in the sense that the countries did not have problems servicing their debts. (Pira & Fuentes 1987).
Later it became evident that extra-regional payments had to have priority in case of current account difficulties to secure a continuous supply of scarce foreign exchange. Apparently, one important reason for the problem was the exclusion of primary products from free trade. This tended to hinder a regional payments equilibrium from establishing itself, since some of the Central American countries, particularly Honduras, could not pay for their intraregional imports only by exporting manufactured goods (cf. Lizano 1989, p. 286).
Forthe small, open economies of Central America, disturbances in their external trade can be very destabilizing. Regional imbalances, in particular, are hard to avoid if stabilization policy is not coordinated between the countries involved, as was the case in the CACM (Lizano 1989, p. 286).
The emerging balance of payments crisis can be analyzed according to a distinction between a "financing phase" and an "adjustment phase" (Bulmer-Thomas 1987b, p. 145). The immediate reasons for the final breakdown of the CACM were, according to Bulmer-Thomas (1987b, p. 89), four: l)A decline in extra-regional exports, leading to a corresponding fall in intraregional trade. As noted above, imports could not be cut as easily as before, due to its heavy contents of capital and intennediate goods. What could be trimmed was, of course, the intra-CACM imports, which consisted mainly of consumer goods. (Before the final collapse of that trade there was a last outburst of activity within the CACM, largely due to Nicaragua's reconstruction efforts after the revolution of 1979. This expansion was shortlived, however, due to the inability of Nicaragua to repay its intra-regional debt.) Public expenditure rose rapidly both just before and right after 1980, due to investment in infrastructure, but also due to increasing military spending (Bulmer-Thomas 1987b, pp. 146-147, Pira & Fuentes 1987). The tax system, however, was old-fashioned and undeveloped. Hence, the expansion of the public sector had to be Imanced mainly by borrowing, an alternative that was facilitated by developing financial markets ll . Despite this, an increasing share of the public borrowing had to be international.
The rising world interest rates exacerbated the external disequilibrium. In the region as a whole, the total foreign public debt grew twenty-five-fold between 1960(Pira & Fuentes 1987. The effect was highly inflationary, although the rates of inflation differed widely between countries. Once the external shocks triggering the crisis had hit the economies, vicious circles tended to develop: economic problems contributed to create social unrest, social unrest ruined prospects for investment, including foreign direct investment, trade had to be cut down, aggravating the economic situation for the region as a whole, and so on (cf. Lizano 1989, p. 289). 2) The at!iustment and stabilization programmes which had to be carriedout after the debt crisis in the early 1980s. The Central American countries atfrrst chose to increase their foreign debt but when the degree of indebtedness became unmanageable they had to sacrifice economic growth (Lizano 1989, p. 291).
This has been called the "financing phase" of the adjustment process, a phase that started in Costa Rica and El Salvador, but which was not initiated in Guatemala until 1982(Bulmer-Thomas 1987b. War and social unrest caused an additional decline in production besides the austerity measures. At the same time demand for consumption goods declined, both as a consequence of the falling real incomes, but also due to deteriorating income distribution (Bulmer-Thomas 1987b, p. 147). In particular, the trend for real wages was essentially negative for the whole 1980s, with the possible exception of Costa Rica. This is illustrated in Figure 11.
3) A series of ad hoc measures of a "beggar-thy-neighbour" nature were imposed to keep the balance of payments under control (restrictions on imports from other CACM members, exchange control delays in import licensing, changing exchange rates, and non-payment of debt in arrears). Atthe same time, great fluctuations in real exchange rates, as well as substantial capital flight (Bulmer-Thomas 1987b, p. 147), aggravated the balance of payments problem.
Payment arrears and increasing protectionism in turn caused creditors (primarily Guatemala and Costa Rica) to restrict their intra-regional exports. A considerable part -but not all-of the problem apparently emanated from the failure of the Sandinista government in Nicaragua to service its intra-regional debt (Bulmer-Thomas 1987b, p. 146).
This was the "adjustment without conditions" phase (Bulmer-Thomas 1987b, pp. 158-160). For details about different trade restrictions imposed after 1982, see Brock (1989). It should be noted, however, that a sizable part of the intra-regional transactions still use the services of the Clearing House (in 1986(in ,45%, according to Brock 1989, although the figure is far from the 100% of the early 1980s.12 The public sector continued to run large deficits. A major reason for this was the collapse of the foreign sector, which heavily affected the taxes that could be collected. Secondly, public expenditure was still high. Necessary restrictiveness in government finance could not be reached, since the organized interest groups in the society were all anxious to protect their positions. (cf. Blomstrom & Lundahl 1989, p. 16). 4) Finally a phase of "adjustment with conditionality" was carried out under the auspices of the IMF (a survey of the programs is given in Bulmer- Thomas 1987b, pp. 152-155). A central part of that policy was a reduction of the public sector deficits, part of which emanated from loss-making government agencies. Increasing the non-trade taxes proved difficult, however, and contributed to social unrest. The emphasis on export promotion was now stronger than before, possibly because of the new Caribbean Basin Initiative, launched in 1984(Bulmer-Thomas 1987b. When all the above problems are summarized, it seems quite clear that the economic and political shocks hitting Central America from about 1979 on were so severe that the problems emerging should not be surprising (cf. Irvin 1988).
They cannot reasonably be blamed on the inherent weakness of the CACM only, even if the region was probably more vulnerable than if the industrialization had been more outward-oriented.

Should the CACM be Revived?
The second half of the 1980s seems to have brought an end to the deterioration of the economic conditions in Central America. This stabilization has taken place after a prolonged period of recession, however, and the economies ofthe countries are not very well prepared to embark on anew, more rapid path of growth. There can be no doubt about the proposition that a new opening up of the Central American economies is a necessary precondition for the take-off of a new phase of development. Whether this process should be based on a revival of the integration model of the CACM or not, is a more ambiguous question. The preceding analysis suggests that the strongly inwardlooking, although regional, approach of the past is not likely to be the optimal model!3. Whatever strategy is chosen, the political obstacles to a serious new start in economic cooperation are now (in November 1991) smaller than in decades, in spite of the fact that problems still remain and most likely will continue to do so for a long time in the future.
The doubts expressed about the viability of the earlier model ofintegration does not mean that a certain amount of import-substitution would not be justified, as even a standard neoclassical trade model with a growing stock of capital can demonstrate (see e.g. Krueger 1977). In spite of everything, it should be remembered that the ratio of intra-regional trade to total trade is still fairly high, and hence the countries are very important trade partners for each other as Figure 6 shows, although intra-regional trade has been diminishing. Moreover, the CACM episode has not necessarily been in vain from an overall point of view, even though the physical output gain is now all but lost. There is reason to stress that the failure of CACM is a relative one in the sense that the expectations placed on the potential outcome of integration were clearly excessive (cf. Lizano 1989, p. 289). In spite of the problems experienced, a revival of a closer cooperation between the economies of Central America would no doubt be justified for several reasons, not least due to possible advantages in international negotiations concerning trade and foreign debt. To abandon the integration scheme in favour of national import substitution would also mean a loss of the potential dynamic gains from integration (e.g. economies of scale).
In spite of everything, CACM contributed to a development that must be considered favourable in several respects. In the long run, it is probably nontangible effects on entrepreneurship and human capital in general that will prove to have been the greatest gain, in spite of the fact that most of the gains in terms of GDP were lost in the 1980s. Also infrastructural projects, development of the education system, and agricultural R&D are natural cases for cooperation,just to mention a few possibilities. One of the biggest lasting benefits of regional cooperation may, as a matteroffact, be enterprises of this type. Finally, it is important to recognize that intra-regional trade provides an opportunity to maintain a more stable economy than in a case where the economies are entirely dependent on world market prices forprimaryproducts. Another point is that a regional market may give firms a chance to cover their fixed costs on an extended home market, so as to enable them to export at marginal costs (Fuentes 1989, p. 106).
The further scope for import substitution at some stage becomes exhausted, however, unless new incentives are provided (cf. Reynolds 1978, pp. 253-254), even if the Common Market could be reconstructed. One option would be to go into production of intermediate and capital goods, a step notoriously difficult for import substituting countries. In any case, the structure of protection would have to be entirely different than that of the early CACM, which, as we noted above, tended to favour production of final consumption goods. The new CET of the CACM, adopted in 1986, actually implies a step in this direction, even if it does not reduce the average effective rate of protection (cf. Saborio 1990, p. 293, Chapman 1987. Another alternative would be a shift toward export-promotion of industrial products (Reynolds 1978, p. 253, Bulmer-Thomas 1979. This may not be easy to achieve, as evidenced by the fact that such a switch has been recommended for decades by now, even if the actual incentives to do so may not have been the best possible. It is hard to avoid the conclusion that this strategy seems to be the only one that can provide markets of a sufficient size in the long run. Considering the sizable external debts (with the possible exceptions of Guatemala and EI Salvador, which have been able to stick to their original payment schedules 14 ) there are no credible alternatives to increasing exports. The industrial base built up during the successful period of the CACM has subsequently deteriorated seriously and, given the high rate of effective protection, would probably not have been very competitive even at that time.
Many, more or less ingenious, proposals on what to do about the import substituting sector versus the need to create stronger incentives for extraregional exports have come forth during the 1980s, including a system with multiple exchange rates (Bulmer-Thomas 1989, pp. 72-77)15. Some substantial fiscal incentives for exports have been tried (Blomstrom & Lundahl 1989, p. 19). It should be noted, however, that many of the efforts to revive the extraregional exports are based on utilization of primary (although "non-traditional") products, such as fish, shrimp, flowers etc. (Blomstrom & Lundahl, 1989, p. 18). Much of the "non-traditional" production cannot be expanded beyond a limit, however, and thus cannot be a final solution to the problem of increasing export production, partly because the production offood cannot be neglected, but also because the most pressing long-run problem appears to be the environmental issue (cf. Blomstrom & Lundahl 1989,pp. 37-39). If not halted in time, environmental decay -particularly erosion -will seriously aggravate the development problem. Also with this in mind, a new effort at industrializing, this time in a context of a more open economy, seems highly recommendable.
Continued industrialization would also be desirable because of its spillover effects on know-how and technology. Some subsidization is likely to be necessary here, but it will not be easy to raise the necessary funding, given the financial situation of the public sectors ofthe five republics, in particular in Honduras and Nicaragua. The sectors with the highest potential comparative advantages may be identified and the anti-export bias (with respect to extraregional exports) should be eliminated. Privatization of state enterprise has been attempted (Blomstrom & Lundahl 1989, p. 17) and should continue in order to relieve the government of the burden of supporting inefficient enterprise activities. Additionally, complementary training and instruction for (potential) businessmen and industrialists is very important (Castillo 1966, p. 101). No real attempt attransferring new technology seems to have been made during the active period of the CACM (or earlier, cf. Castillo 1966, p. 161).
Ifpossible, it may be useful for Central America to applyformembership in a larger free trade area, such as the one being currently negotiated between the U.S., Canada, and Mexico. In fact, some steps have been taken in order to enlarge a revitalized community to Panama, and to achieve some sort of trade agreements with Mexico, Chile, and Venezuela (The Economist, August 24, 1991). This, at least, would alleviate one very real hindrance to export-oriented industrialization: the protectionism ofthe prospective buyer countries. President Reagan's Caribbean Basin Initiative already provides the Central American countries with free access for certain industrial products to the American market during a limited period of time (twelve years). Furthermore, the initiative provides American -and recently also Asian -firms with incentives for direct investment in the region (cf. Blomstrom &Lundahl 1989, p. 19, Central America Report, May 26 andJune 2,1989)16.
The outstanding debt, on both intra-regional and extra-regional levels, is one of the most significant obstacles to rapid development in Central America. Additional requirements fora revival ofthe regional cooperation are that the mutual debt arrears must be settled (this was not provided for by the IMF -sponsored programs in the early 1980s), and that exchange rate uncertainty must be reduced. This means that the foreign exchange rates have to be set at realistic values, which may imply substantial changes from the present rates. This should not be overwhelmingly difficult, due to inflation rates that generally (i.e. with the notable exception of Nicaragua) are under control, at least by a Latin American standard. In the longer perspective it is characteristic of Central America that the movements in real exchange rates have been comparatively modest (cf. Caceres & Irvin 1989, p. 176). The CACM countries will undoubtedly also need a substantial injection of foreign aid and other reviving measures.
The social structure in Central America has to be reformed before real sustainable progress can be made. Even if we disregard the humanitarian aspect, it is very likely that a redistribution of income only would widen the domestic markets in the region considerably, thus providing a broader basis for further industrialization. Efforts to develop social conditions and institutions along with the economy should, at least, be made (cf. Lizano 1989, p. 296). No regional cooperation can be a substitute for the need for social and institutional reforms, not least in the countryside (Castillo 1966, p. 113). This includes the institutions of the CACM, which need to be strenghtened and modernized as well, in order to secure a more equitable distribution of the benefits and an improved co-ordination of national economic policies. A major social challenge is the need to equalize the distribution of income. This is no doubt a very difficult task, and probably one that, once more, would put heavy pressure on government finances. The share of the public sector in the economies is, however, fairly modest, due to generally conservative fiscal policies (except for Honduras and Nicaragua during the last few years). The principal target for these measures must be the rural population, and the core ofthe reform must be changes in the land tenure system. An encouraging fact is that the social aspect is more and more recognized, (cf. e.g. Central America Report 2 June 1989), although the way to a more egalitarian society may still be long. NOTES conclusions about terms of trade tended to depend heavily on the time period chosen (or goods included) is valid; for example, if we start in 1920, it is very hard indeed to [md any significant trend. (See Bulmer-Thomas 1 987a, p. 334) 4 A somewhat different view is given by Bulmer-Thomas (1977). He emphasizes the relative lack of bottlenecks in the food market as a consequence of three factors: price support policies toward foodstuffs, the emergence of an intra-regional market, evening out temporary regional shortages, and the external balance of the Central American countries, which permitted them to import food, without having to consider a "foreign exchange constraint". A similar view is taken by Irvin (1988). Cappi etal. (1978, p. 326) conclude that " ... CA CM is effectively self-sufficient intheproductionojjoodgrains ... ". 5 In fact this was one of the reasons for the trade restrictions imposed in the 1980s: if a country insists on sticking to a fixed exchange rate it has to resort to other means of keeping its current account under control if problems emerge. 6 Willmore (1976, p. 398) gives the following figures for the unweighted average nominal tariffs in 1967: nondurable consumer goods 122% (68% in 1959), industrial inputs 35% (31 %) and capital goods 11 % (9%). The average tariff for all products was 50% (34%). 7 The figures for foreign trade may not be very reliable. Here the trends are much more interesting than the absolute values. 8 The situation varies a great deal between countries, however. While Costa Rica and Nicaragua are among the most indubted countries in Latin America, Guatemala and El Salvador are among the least indebted ones, Honduras taking an intermediate position. 9 The structure of the CET did enable some extemal trade creation, however, as the increases in the effective rates of protection led to increased demand for extraregional imports of intermediate inputs and capital goods (cf. Willmore 1976). 10 Distorted factor prices have another drawback as well: even at world market prices the optimal industrial structure may not be the one the relative availability of factors of production would suggest (cf. Bulmer-Thomas 1979) 11 In fact, the CACM countries experienced a fiscal crisis already in the 1960s, when government revenues declined as a result of the liberalization of trade and the establishment of the CET. This resulted in the San Jose Protocol of 1967, which allowed the members to impose a surcharge 0[30% on the CET. Nevertheless, this measure only increased the incentives for domestic production and the bias against exports of manufacturing, without solving the deficit problem (Bulmer-Thomas 1987a, p. 181). 12 As a matter offact, the figure exceeded 100%, suggesting that payments other than those for trade flows were handled by the CCC (cf. Brock 1989, p. 315). 13 Even if integration is abandoned as a strategy for development, there are several fields on which more cooperation would be desirable. Lizano (1989, p. 302) mentions the following: agriculture, natural resources, project formulation, exchange of information, research, and prevention ofnatural disasters. 14 In terms of actual disbursements Guatemala and El Salvador fare much worse, however, as a result of the outcome of the negotiations between these countries and their creditors (Pira & Fuentes 1987). 15 While it can be demonstrated that such a system could be effective in combining the import substitution and export promoting strategies, we feel somewhat sceptical about this suggestion, partly due to potential unproductive rent-seeking (cf. Bhagwati 1982) and because such a system is considered unacceptable by the IMF. The latter consideration is quite relevant for countries with an external indebtedness of the magnitude exhibited by the Central American economies. 16 The value of the CBI is questionable, however, due to its political selectivity and to the fact that most ofthe Central American products already enjoyed duty-free access before.