Resuming growth in latin america: short and long term policies

The authors of this paper assert that the paralysis of the state generated by the crises of the 1970s and 1980s deprived the economies of the region of an important lever to resume and sustain growth. They thus maintain that to overcome stagnation it will be necessary to reconstruct the state’s capacity to implement pro-growth policies. Following Keynes and Kalecki’s ideas, but also classical development economists, the authors argue, first, that short-term macroeconomic policies, to reduce unemployment and to increase the degree of capacity utilization, should be used to promote the generation of profits to firms and to wake up entrepreneurs’ animal spirits. Short-term expansionary policies should be coupled with measures to improve competitiveness and avoid balance of payments problems. They also claim that alternatives to the liberal programme will fail unless a pro-growth strategy is adopted which includes both shortand long-term policies. They thus propose that long-term policies must complete the package, signaling: a) sustained increases of effective demand in the future; and b) investment priorities to ensure that capacities will be created in strategic sectors and branches of the economy.

Latin American economies have performed poorly at best at least since the early 1980s, except for one very important, but rather solitary, accomplishment, the control of inflation.Control of inflation, however, was not followed by resumption of growth, as it was expected.Liberal opinion tends to attribute this poor performance to the reluctant adherence to liberalizing or market-friendly reforms, and more particularly to the still strong presence of the state in the economy, as an agent or as the enforcer of allegedly excessive regulation.
The opposite view, to which the authors of this paper largely subscribe, sustains that it was rather the paralysis of the state generated by the crises of the 1970s and 1980s that deprived the economies of the region of an important lever to resume and sustain growth.The predominance of liberal economic policies all over Latin America in the 1990s served only to make stagnation trends stronger.Brazil and Mexico are just two among many examples that could be named of economies where economic performance deteriorated drastically in the liberal era (the 1990s and 2000s) in comparison with the rates witnessed from the late 1940s to the 1970s.
It is true that Chile, which is frequently used as a show-case for liberal reforms, has achieved above-average growth rates from the mid-1980s onwards.But we must not forget that in the period when it rigidly implemented liberal policies it suffered two deep crises (1975-76 and 1982-83).Moreover, it resumed growth after the latter crisis only because it implemented a more pragmatic and realistic set of policies.Finally, its more recent good performance has had very little to do with adherence to liberal strategies; but it is rather due to a combination of economic pragmatism and good luck (Ffrench-Davis, 2004).Argentina is another interesting case since it reached (and sustained so far) Asian rates of growth after breaking up with the Washington Consensus.
The main hypothesis underlying the arguments to be presented in this paper is that the failure to engineer sustained growth is rooted in the paralysis of the state that followed the crises of the late 1970s.If this hypothesis is correct, then, to overcome stagnation it will be necessary to reconstruct the state's capacity to implement pro-growth policies.
The need to reactivate the state is not in itself, of course, a new or revolutionary proposition.However, most critics of the liberal view have tended to assume one of two positions.Some believe that all that is needed is to implement expansionary policies that stimulate domestic demand to revive employment and output, "to restart the engine", so to speak, that the economy will then move on and begin growing powered solely by private entrepreneurship.Others, on the other hand, seem to believe that there are no alternatives to orthodox macroeconomic short-term policies so that non-liberal critics should focus only on long-term possibilities, policies and instruments, such as the formulation of industrial policies.At the risk of caricaturing both groups, one could say that the first group believes that non-orthodox short-term macroeconomics policies should do the trick, while the latter believes that the macroeconomic battle is lost anyway, so that salvation resides in choosing the right long-term microeconomic policies.
Following Keynes and Kalecki's ideas, but also classical development economists (Prebisch (1949), Furtado (1953), Mandelbaum (1945)), we propose in this paper a different view.We argue, first, that in a modern capitalist economy, a monetary economy, short-term macroeconomic policies, to reduce unemploy-ment and to increase the degree of capacity utilization, should be used to promote the generation of profits to firms and to wake up entrepreneurs' animal spirits.Moreover, and this is equally important, short-term expansionary policies should be coupled with measures to improve competitiveness and avoid balance of payments problems.We also argue, in second place, that alternatives to the liberal programme will fail unless a pro-growth strategy is adopted which includes both short-and long-term policies.We thus propose that long-term policies must complete the package, ensuring sustained increases of effective demand in the future; and signaling investment priorities to guarantee that capacities will be created in strategic sectors and branches of the economy.In developing economies, where resources are not abundant, production chains are incomplete, and competition may be strong, industrial policies are necessary to coordinate private investment plans, thereby ensuring that current and foreseeable bottlenecks will be taken care of, and giving businessmen some measure of safety as to the feasibility of their plans.
To argue the need for active short-and long-term policies, we begin, in section 1, by setting up our theoretical approach with a brief presentation of the relevant Keynes/Kalecki framework connecting aggregate demand and aggregate profits, complemented by a discussion of the need for long-term growth-promoting policies.Section 2 explores the possibility of adopting expansionary short-term policies in Latin America.Section 3 discusses the necessary elements to define a supporting long-term investment strategy.Section 4 illustrates the points raised in the preceding sections with a simulation exercise developed for the Mexican economy.Section 5 concludes the paper.

CYCLICAL RECOVERY AND SUSTAINED GROWTH
In the perspective of this paper, involuntary unemployment (and idle capacity) emerges when aggregate demand is insufficient to absorb the full employment level of output.For Keynes, as for Kalecki, this sort of situation in fact tends to perpetuate itself since, everything else remaining the same, if actual -and expected--demand is low, few jobs are offered, so few people will actually be able to buy consumption goods, confirming the initial expectation of low demand.Furthermore, as Minsky (1982) reminded us, when aggregate demand is low, profits are low and firms may not be able to honor the debts they incurred in the past without selling assets or issuing in more debt.Under these circumstances, even if realized or expected demand for some entrepreneurs' output improves, the situation may not get better if firms are enmeshed in debt.The existence of debt liabilities means that for firms even transitory falls in demand may have lasting adverse effects.
In the Keynes/Kalecki approach, therefore, short-term policies to maintain appropriate levels of aggregate demand are essential to bring about high profits and keep firms financially healthy.If a recession has already started, even more vigorous policies may be necessary to increase demand, thereby allowing firms' revenues to rise, and generating profits.1 Supply-side measures leading to cheaper raw materials or to lower interest rates, though probably helpful, may not be enough if firms do not sell their cheaper output.Lower interest rates may help firms to attenuate the impact of the fall in profits on their ability to honor debts, but should hardly induce them to increase their financial exposure by borrowing to increase production if demand remains subdued. 2 Now, growing profits generated by increasing aggregate demand should improve the lot of firms through a number of ways.First, they provide them with cash flows, making it possible to service their debts.Secondly, the increase in internally generated funds also allows them to risk increasing output without exposing them to additional financial risks.Thirdly, as Keynes argued a few times, there is no better way of improving expectations than the actual experience of earning profits. 3The increase in profits can thus contribute to improve the state of confidence in the economy causing entrepreneurs to engage in increasing production as well as generating internal funds to allow them to do it.
It should be obvious that a recovery of aggregate demand has to originate in those of its elements that are autonomous with respect to current income.Private domestic expenditures may not be the most promising candidates in this particular.It is true that a significant fraction of consumption nowadays is actually autonomous with respect to current income, being financed by consumer credit, as it is the case of higher-value durable consumption goods.However, we do not think that a consumer credit-led recovery is the most adequate alternative for our countries.
Private investment will hardly be promising: firms do not have any reason to invest if they already have idle capacity and even less if they are facing problems to honor their debt-service commitments.
Under these circumstances, the two avenues left to promote recovery are net exports and what Kalecki (1933) called "domestic exports", the increase of sales to the government.Expanding net exports has the advantage that it prevents or at-tenuates balance-of-payments' current account problems caused by the increase in imports that usually accompanies an economic recovery.The means to stimulate export expansion and import substitution will be addressed later in this paper.
The other possibility is the use of fiscal policies.The latter can push an economy onto the path to recovery.Contrary to widely shared notions, though, Keynes and Kalecki argued that what is needed to revive the economy are not fiscal deficits, but increased government spending.Kalecki in fact supported financing government expenditure with taxes on profits; which would contribute to demand expansion combined with a change in income distribution in favor of labor. 4Keynes, on the other hand, saw the generation of fiscal deficits a kind of last resort instrument to be adopted when nothing else worked to fight a recession.Keynes believed that public confidence that the government would act if necessary to fight a recession should be guarantee enough to motivate entrepreneurs to maintain adequate expectations.Spending policies did not imply generating deficits because of the income multiplier that would expand income and, thus, tax revenues when government made its expenditures.If a deficit was generated, its final impact on the economy would depend, to some extent, on how it was financed and the public's liquidity preference that would define the terms on which debt securities could be absorbed in private portfolios. 5Therefore, for countries with significant structural fiscal deficits, the appeal to political fiscal may demand some previous reconstruction of the ability of governments to control their own fiscal situation.Now, in the case of developing countries, public spending should emphasize in the first place public investment, since the infrastructure provided by the state is fundamental in the process of economic growth.Indeed, as classical development economists long ago made it clear, in developing economies the problem is not only to maintain proper levels of aggregate demand, but also to solve short-term bottlenecks that hamper supply, and to expand productive capacity.In the second place, given that developing countries in Latin America also exhibit appalling income concentration indices, with lower income groups living in very bad conditions, it is likely that income transfers to the poor may also have to be increased if for no other reason to show that development is a process that benefits the whole society and not only privileged groups. 6n addition, this fiscal push should be accompanied by an accommodating monetary policy that would prevent an increase in interest rates that could create more problems for the private sector.Increases in aggregate demand will raise profits but it should not be inflationary if there is idle capacity and involuntary unemployment. 7dopting expansive macropolicies should be able to accelerate short-term growth in Latin American economies beyond the mediocre levels they have reached in the not so recent past. 8However, they do not by themselves create conditions for a permanent acceleration of growth after capacity utilization reaches its limits, whatever elastic these limits may be in the short-term.They lead to increased output and employment and increased profits, but this may not be enough to induce investments; or to induce the right type of investment.Therefore, government policies should signal investment opportunities to entrepreneurs at the same time in which they create conditions to facilitate the realization of investments.Active promotion of exports and import substitution, for instance, can at least in part fulfill this role.Industrial policies to support investments and to stimulate innovation and increasing productivity are another instrument.9Developing economies exhibit gaps in their productive structure.These gaps may end up causing some investment strategies to become non-viable.The role of industrial policies is to signal investment priorities, and to coordinate private decisions in order to anticipate or attenuate the uncertainties created by the existence of these gaps.Businessmen can go ahead with their investment plans when they are sure that their needs in terms of materials and equipment will be provided for.In any case, these policies are not alternatives to short term recovery policies, but their complements.To affront the uncertainties of long-term investment is at the very least made much easier if firms are already enjoying healthy cash flows as a result of short-term growth, demand expectations are already favorable and animal spirits are awake.
Therefore, if a recovery is to become an increasing trend, to use Kalecki's terms, businessmen have to be induced to invest beyond the mere replacement of depreciated capital, and this means that specific measures signaling that future de- 7 In section 3 we debate in some detail the notion of idle capacity. 8Of course, liberal economists argue that activist states scare away "investors" (it is never clear to which class of "investors" one refers here, since most of the time it seems that the argument relates to financial investors, not really investors in productive facilities).For them a small and balanced government stimulates businessmen to run risks.Through some obscure economic logic, it is argued that macroeconomic equilibrium would stimulate private investment.In contrast, the argument presented here is based on the idea that investment is a disequilibrium phenomenon since it is induced by an expected disequilibrium between supply and demand large and durable enough to justify building new capacity.Thus, the role of macroeconomic policy is precisely to generate the correct type of disequilibrium.Of course, this should not be construed as giving a blank check to government to spend in whatever ways.There are perverse situations to be avoided like stimulating further an economy already in full employment.The solution, however, is not the elimination of the government as an economic factor, but in defining efficient ways to make good its potentialities.mands will justify these investments are essential.Also essential are the measures to make feasible the funding of these investments.In the developing economies of Latin America this means mainly to make sure firms will be able to generate the necessary funds internally, but it also requires that new financing channels are open so that firms can invest beyond the limits set by their own funds.
In the remaining sections, we will propose a strategy whereby these principles could be translated into actual policies in the case of the largest Latin American economies.

Can a higher growth rate be achieved in latin america in the short-term?
There is abundant evidence of the short-term potential for accelerating growth in Latin American countries.On the one hand, available information about unemployment rates and capacity utilization shows that there are large reserves of productive resources kept idle in these economies.On the other hand, most economies in the region react very quickly when expansive policies are implemented or perspectives of wider markets are created, showing no lack of entrepreneurial disposition to take advantage of market opportunities.
Even though the region has exhibited high growth potential as well as high sensitivity to increases in aggregate demand, Latin American economies have been prevented from actualizing their potential by the generalized adoption of contractionary macroeconomic policies.Only those countries who were able to shake off the constraints imposed on them by Washington Consensus-inspired policy strategies could enjoy their potentialities.As can be seen from Tables 1 and 2, Argentina after the 2001/2 crisis and Brazil since 1994 illustrate both how fast an economy can growth in response to expansionary policies (the former) and how an economy can waste its potential in semi-stagnation (the latter).We begin by examining the availability of idle resources.It is an empirical fact that labor unemployment is widespread (cf.Table 3).Open unemployment figures do not tell the whole story, because precarious or informal employment is huge.The estimated figures of informal employment as a percentage of the labor force for selected countries were as follows.Argentina, 42.5 in 2002; Brazil, 46.2 in  2001; Mexico, 47.2 in 2002; Uruguay, 45.7 in 2002; Venezuela, 56.5 in 2002.The figure was high even in Chile (32.5, in 2000), where the rate of growth has been well above the regional average for about more than two decades. 10The existence of unused resources is also shown in capacity utilization data.Figures on this point are harder to come by, but, according to Banco de Mexico's surveys, in 2003 firms used only about 72 percent of their capacity, and that rate never exceeded 75 percent in the 1996-2003 period.In Brazil, even in 2004, when the highest rate of GDP growth in recent years was recorded, capacity utilization reached only 82,7%.In Argentina, the degree of utilization of capacity in the manufacturing industry has remained below 72% between 2004 and 2006; in spite of the fast growth achieved by this country.In fact, if anything, available information might probably underestimate these economies' growth potential.It is well known the discouraging impact that prolonged unemployment exerts on workers leading many of them to actually withdraw from the labor market (and the labor unemployment statistics).On the other hand, capacity utilization is not, in fact, a material constraint defined by engineering calculations.The degree of utilization is estimated by businessmen, who use as benchmark the level of production they reckon as optimal.It is only natural that these estimates themselves are influenced by expectations and by the state of confidence.When demand is brisk and profit expectations are high, businessmen naturally take a higher level of production as their benchmark.11It is not only because in a prosperous environment businessmen' animal spirits is stimulated, but they are also stimulated to explore, for instance, x-efficiency factors, that is, to increase their operational efficiency.Thus, capacity utilization measures themselves show some "elasticity" with respect to the growth of demand.It has been in fact frequently observed that in fast-growing economies capacity utilization grow less rapidly than GDP, because businessmen keep updating their estimates of their maximum possible output. 12ainstream economists usually fail to see that unutilized capital equipment can be important for growth.They implicitly assume that firms will keep idle only that equipment which has no economic value.This may come from the notion that in a general equilibrium model with perfect competition, any economy will produce on their production possibilities frontier.However, even neoclassical economic theory, when analyzing so-called "internal distortions", draws a different conclusionnamely, when social and market prices and costs do not coincide, the economy will not reach their Pareto optimum.Thus it will not optimally use factors of production.In this context, we could point to the so-called domestic cost of resources, which is a useful indicator to assess whether engaging in domestic production is worthwhile, or, to put it differently, whether idle capacity has economic value.The procedure consists in taking, for any activity, domestic value added at shadow prices, dividing it by the shadow price of foreign exchange, and comparing the result to value added at world prices.If that ratio is below one, the activity is efficient; the sector has comparative advantage and macroeconomic efficiency would advise to use those resources. 13Huerta and López (2006) employed this methodology in a recent study of Mexico's manufacturing sector and concluded that producing with idle equipment would have been economically worthwhile in 46 sub-sectors in 1990 and in 45 in 1996 (out of 49 sub-sectors).
Anyway, better use of currently existing capital equipment may allow fast growth to take place.Growth recovery is important in itself, but it also essential to stimulate investments necessary to increase capacity in the future.Argentina's recent economic experience illustrates the point.This is shown by the following graph (adapted from Damill, 2006), where we compare the behavior of the index of economic activity, and the index of capacity utilization in manufacturing (both indices are set to 100 for 2002.4) In Argentina, GDP growth averaged about 9 percent between 2003 and the second quarter of 2006.Even though the productive capacity expanded, it would appear that the fast economic recovery in this country was possible mostly thanks to a greater utilization of the previously idle capital equipment.In fact, as shown in the graph the capacity utilization index grew faster than the economic activity index until 2004, when investment started to grow fast.Most likely it was the recovery that stimulated the investment activity after that date.It should be noted that the rate of utilization of capacity was still at only 72% at the end of 2005, in spite of the fast economic growth achieved.
There is no evidence, thus, that the large Latin American economies are producing at their potential output so that any growth acceleration would necessarily be inflationary.Given the overall abundance of labor in the region, the index of idle capacity would give us a rough measure of how much room there is for 13 For any i activity we express the Domestic Cost of Resources DCR as: where and w denote the shadow price of capital and labor, L and K the amount of labor and capital required, respectively; Y and S the gross value of production and of direct and indirect required inputs, also valued at shadow prices, respectively.production to rise in each country.Occasional bottlenecks in one or another sector could be attenuated with imports.But how could these countries reach higher levels of capacity utilization?

Growth-oriented short-term policies
The most obvious requirement of a short-run economic expansionary strategy is to increase aggregate demand.On this point, we recall again that Keynes  (1933, 1936) and Kalecki (1944), strongly recommended expansionary fiscal and monetary policies. 14The European economies in effect successfully carried out expansionary policies in the immediate post war period.One should beware of mechanically transposing these experiences to the Latin American case though.The institutional underpinnings of Keynes' and Kalecki's arguments, as well as the context within which the European experience unfolded, were different from what we see in Latin America today.In particular, Europe had protected domestic markets, and strictly controlled capital movements and this was the context assumed by both Keynes and Kalecki.Besides, all the advanced capitalist countries were following expansionary policies at that time, which allowed the mutual strengthening of growth stimuli.
Latin American economies should certainly learn from the European (and other countries') experience, but one cannot forget that, being still developing economies, not only demand-side, but also supply-side measures are fundamental to achieve higher growth rates.Thus, a second requirement of economic expansion is that growth of domestic production should be accompanied by strong efforts to improve competitiveness of domestic producers.In the short term, this practically means management of the exchange rate.In the longer term, there is also the need for supporting policies, and most notably by industrial policies and in general structural policies.
Theoretically, a large enough currency depreciation would make locally made goods both profitable and competitive.In fact, even production with old and highcost equipment which today remains idle may become competitive depending of the extent of currency depreciation.However, a large depreciation, unless it is accompanied by supportive measures, also creates some risks; particularly in economies that are very open to trade and finance.First, the price hike it may trigger can give rise to adverse expectations about the future stability of prices and the exchange rate.This by itself may lead to capital flight and threaten the balance of payments.Second, depreciation brings about a shift against wages and unleashes inflationary pressures.If money wages fully adjust to prices, a large depreciation of the domestic currency will provoke a wage-price spiral. 15If they do not fully adjust, real wages and consumption per worker will fall; and obviously a wage fall is something that a progressive government could not accept.Currency depreciation has an additional negative impact on income distribution, because it may give extra profits (but not necessarily higher unit wages) to sectors that were already competitive and highly profitable to begin with.This is particularly the case of exporting sectors based on natural resources. 16Finally, depreciation increases both the debt ratio of firms and of banks, when they are indebted in foreign currency; and the supply price of imported capital goods.Thus it may weaken the lending capacity of banks, and discourage private investment in fixed capital.
It follows that currency depreciation, adopted in isolation, while it does stimulate exports and import substitution, may harm consumption and investment, even as it worsens income distribution.Some evidence shows that its negative effects on internal demand may be larger than the improvement in trade balance brought about by the depreciation. 17e it as it may, a competitive exchange rate seems indispensable, and in any growth resumption strategy, competitiveness of domestic production must be given a boost (Bresser-Pereira and Nakano, 2002).Argentina's recent experience suggests that under certain circumstances the harmful consequences of currency depreciation can be reduced, while its beneficial effects can be fully reaped.
To achieve the gains of currency depreciation even as its negative consequences are minimized, we propose a "compensated devaluation". 18That is, the currency depreciation must come hand in hand with complementary measures.In the first place, the government must ensure that the real purchasing power of workers does not decline; subsidizing them for any loss in their real wage. 19Secondly, measures must be taken that compensate firms, at least partially, for the effects of currency depreciation on debt servicing and on the cost of imported inputs.Thirdly, taxes must be levied on extra profits earned by exporting firms that prior to the currency depreciation already fetched high profits to prevent the "dutch disease". 20Anyway, the most important compensation for capitalists and for workers for any harm ensuing from devaluation is output and employment expansion.
Both the currency depreciation and the complementary measures can take many forms.For example, the depreciation may involve a dual or multiple exchange-rate system.Alternatively, depreciation could go together with subsidies to domestic production, provided that firms pledge to pass on lower costs to prices and measures are taken to ensure that this is actually the case.A country could get an analogous result with a currency depreciation compensated with a discount in the Value Added Tax for basic consumption goods, or compensated with money transferences to lower-income groups.We cannot go into the details here.We just want to point out that before the adoption of the liberal credo, Latin American nations accumulated a rich and varied experience in this area that is very much worth studying. 21he third requirement of a short run strategy is selective credit expansion.To be able to respond to demand, firms must have the financial means to invest in extra working capital at low interest rates.Thus, well-targeted short-run selective credit policies are fundamental for a growth resumption strategy.Of course, this entails adopting the notion that the objective of monetary policy is safeguarding macroeconomic stability rather than just price stability.There is in fact no reason to expect that, when output is still elastic, increases in the supply of credit are inflationary.If, in addition, measures are taken to make sure that credit will be directed at priority sectors, including the production of tradables for export and import substitution, the impact on prices and on the trade balance should be greatly attenuated.
a fourth requirement is the ability to regulate capital movements.If it wants to use the exchange-rate as an instrument, as we think it should, the government must minimize speculative-or interest rate arbitrage-induced currency swings.To reach this objective, the Central Bank should quickly intervene whenever undesired variations of the real exchange rate take place.Also, the monetary authorities should be open to impose controls over capital movements in case they are necessary (Bhaduri and Matzner, 1990; Steindl, 1990; Bresser-Pereira and Nakano  2002; Frenkel, 2006; Sicsú and Ferrari Filho, 2006).
Capital controls serve many objectives in the picture we are drawing, the most important of which is to prevent sudden changes in market sentiment, to use Michel Camdessus' expression, from eroding the influence of the monetary authorities in the stabilization of exchange rates and in keeping interest rates low.Under capital account liberalization, as capital movements in and out of the coun-try are free, exchange rates are much more influenced by conditions in the financial markets than by trade considerations and interest rates have to be kept higher than adequate to the needs of the economy in order to avoid capital flight.
Capital controls should be considered under three main headings.Financial investments of non-residents should be subjected to entry requirements, as, for instance, those so successfully implemented by Chile in the 1980s and 1990s.Investments abroad by residents, in contrast, should be restricted by exit limitations.Finally, foreign indebtedness of domestic private firms should be controlled and limited to exporting businesses, which enjoy a "natural" hedge against adverse changes in exchange rates, since their revenues are linked to foreign currencies.These three measures should be enough to provide the government the necessary control over exchange rates to use them as an instrument of export promotion and import substitution and room to set domestic interest rates according to growth objectives.

LONG-TERM ELEMENTS OF SUSTAINED GROWTH POLICIES
Let us turn now to the medium and long-run factors to be considered when formulating an alternative growth strategy to that proposed by liberals.We concentrate our attention on how Latin American economies can achieve high rates of growth of productive capacity, and how rapid growth can be made compatible with external equilibrium.
Investment plays a central role in the process of growth of productive capacity.A rise in the investment ratio is essential to accelerate the growth rate of output.The pioneers of development economics (Mandelbaum, 1945; Prebisch,  1952; Furtado, 1953; Kalecki, 1969) stressed two desirable features of investments appropriate to countries where unemployment is widespread: capital-output ratio should be low, and the life-span of the capital equipment should be extended as far as possible.Under these circumstances, high output growth rates could be achieved even with modest investment rates, or, in other words, without sacrificing too much present consumption. 22s this still a sensible recommendation?We think it is.But we also qualify their proposal.Indeed, extending the life-span of the capital equipment and investing in equipment characterized by low capital intensity may negatively affect the rate of growth of labor productivity.
A strategy entailing stagnation, or low growth, of labor productivity, will soon be exhausted.First of all, once full employment has been reached, the overall rate of growth of output cannot exceed the rate of growth of employment, plus the average rate of growth of labor productivity.If the latter is not high, output growth will be insufficient to achieve a high level of output per head.On the other hand, there is the foreign trade barrier.Let us assume that output growth shoots up under such a strategy.We know that to keep the trade balance in check, the rate of growth of exports must be accelerated, or the import-elasticity of income should come down, or both. 23Since Latin American economies must grow faster than their trading partners, competitiveness of home production must shoot up.To a certain extent the latter can be improved if the country subsidizes labor costs to bring them closer to their opportunity cost. 24A competitive exchange rate, as well as ample access to credit, would also help.However, subsidies cannot last for ever, and no country could persistently devalue its currency without generating adverse side effects or inducing retaliation.Therefore, a fast-growing country must also carry out efforts to foster technical progress, to change the export-mix in favor of goods for which world demand grows fast, and to sustain increasing levels of labor productivity.This is why we would rather recommend what we call a strategy to "walk on both legs"; to use an expression that in the past had earned a certain reputation.
Firstly, in those sectors that are somewhat protected from foreign competition (or in which the country has important natural comparative advantages), one should seek to extend the lifetime of their capital equipment, and to use labor-intensive techniques of production; provided of course that these techniques are feasible.This should be also accompanied with efforts to diffuse technical progress to these "protected" sectors, so as to upgrade progressively their technical and managerial capacities.However in those sectors that are more exposed to competition, the adoption of more advanced (and probably more highly capital-intensive) techniques will be required.
secondly, Latin American countries should make efforts to change their export and their import mixes.This entails a careful selection of investment projects.These should be focused on the one hand towards future possibilities of import substitution, in modern lines of production which currently satisfy their requirements with imports; and on the other hand towards international demand niches, where they can construct new competitive advantages.
Thirdly, they should foster autonomous technical progress.The latter is especially important and it entails two requirements.On the one hand, there are the formal and the informal educational systems.There are also learning-by-doing effects, and in this regard we recall that there are some particular industries the workings of which appear to stimulate technological progress.Technological development is held back when these technology-generating industries are absent, 23 Thirlwall (1979) expressed this notion, which has come to be known as the "Thirlwall's Law", as follows: y x x = π where y X is the rate of growth compatible with external equilibrium, x is the rate of growth of exports, and π is the income elasticity of imports.In parentheses, the same notion had been proposed earlier by Prebisch (1951 [1982]).and the country runs the risk of losing the basis for autonomous technological development.Often, it takes time to put these industries in place.This is why it is necessary to start them in the early stages of industrialization.
Accordingly, in our proposal to "walk on both legs", we suggest that Latin American economies should also channel a share of their new investment towards the promotion of innovations and the development of a few very modern industries, even if they are capital-intensive and thus require considerable capital investment.Why do we emphasize here the word few?Because saving potential is not high in developing countries, even if all resources were fully utilized; and professional and technically skilled labor is scarce.This limits their capacity to establish technologically advanced industries.

A SIMULATION EXERCISE WITH MEXICAN DATA
To illustrate to the previous arguments, we carried out simulations of alternative policies, using the example of Mexico, for the 2006-2025 period.We will first abstract from the trade balance aspects, and in a first step we will consider three scenarios.The common assumptions of the first two of them are: The two first exercises differ in that: e) In scenario A, we assume a 15-years life-span for investment in machinery and equipment 40 years for nonresidential construction, and 50 years for residential construction. 25Besides, the degree of use of the capital equipment, and the technical capital-output ratio, remain constant.f) In scenario B, we assume the degree of use of productive capacity increases step by step until it reaches 95 percent.26Also, the life span of capital equipment grows, to 60 years for residential construction, 58 years for nonresidential construction, and 20 years for machinery and equipment.Finally, we assume a gradual decline in the capital-output ratio, until it arrives at a value of 2.68 in 2014, and stays constant afterwards.
In Graph 2 we can view the time-pattern evolution for production and consumption under the two scenarios.

GRAPH 2
We can see that the long-run evolution of production and consumption are much better under scenario B. The difference is more patent during the first years, because then the measures recommended have a greater importance. 27Now, we would not claim that scenario B is a realistic one.A growth rate of output of about 13 percent in 2006 and 10 percent in 2007 would probably be beyond reach; and most likely will bring about strong inflationary pressures.But the exercise is useful to demonstrate the potential that can be tapped with a better use of resources.
The difference between the two scenarios narrows after the first years, since the pace of investment sets up a limit to growth, similar under the two scenarios.How would things change if the growth rate of investment were higher?This is our scenario C, where we assume that investment growth 6% annually.We can see the time evolution the growth rate of consumption for scenarios B and C in Graph 3 below. 27In particular, the depreciation coefficient is nil during the first years.

2025.D2
2025.D3  We note that in the latter scenario, the short-run evolution of consumption worsens with respect to scenario B, though not by much.From 2011 onwards, however, the pace of consumption is better under scenario C. Furthermore, the high rate of growth of consumption is achieved without large investments.In fact, the rate of investment would drop from 21% in 2006, to 16% in 2025.However, Mexico cannot maintain such a high growth rate of consumption or production for too long.This is because the labor reserve would be exhausted in year 2020.From that date onwards the growth rate of consumption and output would fall to 3%, because (under our assumptions) the growth rate of labor productivity in the formal sector and the growth rate of the economically active population are 1.5% annually each.
Let us therefore consider a new scenario D. We assume here a higher rate of growth of investment, of 8% per year; which would take the investment ratio to 25% of GDP in 2025.Also, we assume that the capital-output ratio is reduced in the "protected sectors".However, we incorporate also a second, "modernizing" leg, to our strategy.Thus, in selected sectors capital-intensive techniques are adopted.Also, new "modern" industries are developed.All this entails that the average capital-output ratio does not fall as in the previous scenario, and remains at its previous level; i.e. 2.75.But as a counterpart, this makes it possible to raise the rate of growth of labor productivity.This idea is illustrated in Graph 4 below.As we can see, in comparison with scenario C, in scenario D the short run evolution of consumption is somewhat affected.However, starting from year 2014 onwards, a higher rate of growth of consumption is achieved under scenario D, because the rate of growth of labor productivity shoots up.This makes it possible to grow at a faster pace.Furthermore, a high growth rate of output and consumption is maintained, thanks to the acceleration of the rate of growth of labor productivity.
We will study now some foreign trade scenarios.Let us assume an average growth rate of output of 6% per year.We will consider the effects of different economic policy alternatives to achieve this rate.In Table 1 we can see the assumptions and results of three foreign-trade scenarios, labelled D1, D2 and D3.In the first scenario (D1), Mexico's foreign trade parameters would be preserved. 28However the long-run growth rate of investment would rise to 4 percent annually.In these circumstances, import growth would greatly exceed export growth.Consequently, the share of the trade deficit in relation to the GDP, which was 1.3% in 2005, would increase to 6%.This path is not a desirable one for Mexico, since such a foreign debt would be dangerous, and would threaten a new crisis similar to the one experienced at the end of 1994.
In the second scenario, authorities would devaluate the peso in order to improve competitiveness and maintain the trade balance in equilibrium in rapid GDP growth conditions.In this scenario, Mexico would maintain the foreign balance (as a share of the GDP) at the current level.However, devaluation leads to accelerated inflation, which would be 20% annually on the average.Furthermore, real wages would drop, and workers would lose more than 90% of their purchasing power (the purchasing power of wages would drop from an index of 100 to an index of 7).This scenario is socially and politically unfair, and furthermore, it is unviable.
In the third scenario, authorities would moderately depreciate the currency.This would be a "compensated devaluation", which will go hand in hand with the 28 We took most of the parameters from Castro, Loría and Mendoza, 1997.specific type of measures recommended in the previous section.These measures would help to modify some of the parameters characterizing the response capacity of domestic producers.In particular, it would be possible to lower the income elasticity of imports, and Mexican firms could supply an increasingly greater proportion of the domestic demand.At the same time, they would be in better conditions to respond to foreign demand (the income elasticity of our exports would increase).In this scenario, growth would be achieved in conditions of external equilibrium, and with moderate inflation (5% annual average).In addition, rapid employment growth would increase the number of working members per family, which would cause the average income of families to increase.In this scenario, however, real wages per worker would drop, and income distribution would worsen.Therefore, authorities would need to compensate for this situation by granting direct and indirect subsidies to workers, in the framework of a more generous social policy.Subsidies would need to be financed by taxes on the profits and income of the wealthy population groups.

CONCLUSION
Economic growth has been mediocre in Latin America for about three decades.The poor performance of the region, however, has not been caused by natural disasters or social unrest and civil strife.Beginning in the 1970s, these economies lost strength as they suffered a series of strong adverse external shocks, such as the two oil shocks of the 1970s and the interest rate shock generated by the Federal Reserve's twist to monetarism in 1979.After the largest countries in the region, Argentina, Brazil and Mexico, went bankrupt in the early 1980s, the protracted process of adjustment that took place was conducted by the IMF under the aegis of liberalism.In various degrees, the whole region pursued the policies known collectively as the Washington Consensus, dismantling the instruments of intervention of the state in the economy, privatizing companies, deregulating economic activities, removing barriers to the international circulation of goods and capital (but not of labor, as it is usually noted).The liberal rhetoric was adopted to advance the idea that unleashing the energies of the private sector should be enough to inaugurate a new era of growth and prosperity unencumbered by the imbalances of the past, manifested by high inflation, balance of payments vulnerabilities, and so on.
The historical record is unambiguously unfavorable to these views.In fact, as we showed in section 3, growth has been low not because these countries are unable to reach higher output, but because liberal macroeconomic policies have kept actual output much below potential output.In fact, the liberal experiment in Latin America has unwittingly endorsed Keynes' and Kalecki's theory of effective demand by showing that misdirected economic policies can depress aggregate demand thereby preventing the economy from achieving and sustaining its potential output.
In this paper we tried to outline an alternative policy strategy that, based on the theory of effective demand and in the classical development theory, would stimulate the economy to grow in the short term, and sustain growth in the longer term.Short-term growth acceleration can be achieved by stimulation of demand through monetary and/or fiscal policies.We argued, however, that imbalances peculiar to the economies of the region make it necessary to take some additional measures to prevent possible adverse effects of these policies on the profile of income distribution, balance of payments, etc.In our view, more important than the listing of possible instruments, though, are the principles that we are proposing.Namely, that the reconstitution of the partnership between the state and the private sector in the economy is indispensable, and that a successful growth strategy in the region has to contain both short-and long-term sets of policies.Long-term policies will fail in the absence of short-term policies favorable to induce growing output, employment and profits.Short-term policies, on the other hand, per se, may not be sufficient to induce the type of investment necessary to enlarge productive capacity.
Still, even these cares may not be enough since, in contrast to Keynes and Kalecki, we are dealing here with developing economies.Although the logic of capitalist, or monetary economies, is the same in developed and developing economies, supply constraints are obviously stronger for the latter.Therefore, to sustain growth in the longer-term a new set of policies may be necessary, including the use of industrial policy instruments.As an illustration of the main points raised in the paper, we present the results of a simulation realized with data from the Mexican economy.
a) 4 percent investment growth rate.b) Utilization of capital equipment of 80 percent in 2005.c) A value of 2.75 for the capital-output technical coefficient.We obtained this value by dividing our guesstimate of the capital effectively used in 2005 by that year's GDP.d) Balanced foreign trade in 2005 and afterwards.
TE C ON SUM PT ION IN C GROWT H RA TE C ON SUM PT ION IN B GRAPH 3 nOTES: a) Growth rates, b) Percent of unit costs, c) Average for the period, e) Pesos per US dollar.Sources: See text.

Table 2
Per capita GPD Annual Rates of Growth (%)

Table 3
Unemployment Rates