During the past decade, Latin American countries enjoyed some of their best economic times in recent memory. Most of the good performance is attributed to unusually favorable international prices, especially for natural-resource-based products and services. Investment was promising, inflation rates were in single digits for most countries in the region, and unemployment reached its structural levels.1 Other macro indicators looked healthy compared with past figures and, more important, poverty was significantly reduced.
However, it seems that these tailwinds are losing strength. Commodity prices have dropped significantly for the past two years and foreign direct investment, which had flourished, is in a long-term decreasing trend. Today, several countries are facing fiscal problems, especially in the Caribbean, where governments are under pressure to maintain or increase social expenditures to weather the storm. Even though several welfare indicators have improved in most Latin American countries, recent events may jeopardize this performance and the issue has grabbed the attention of local and international policymakers.
One of the main characteristics of the region’s countries is their disproportionate dependence on a highly concentrated bundle of products, most of them natural resources. This dependency explains their recent success, but also their fragility. A natural resources bonanza has kept countries from developing more valued-added sectors that could have buffered economies in lean times. Yet what is most striking in the region’s long-term growth figures is an evident stagnation of total factor productivity (TFP) in almost every country; indeed, it has been on a decreasing trend, and not only in recent years.
This chapter argues that intimately linked issues such as export and output concentration and stagnant productivity need more attention, especially considering the potentially hard times coming in which, as always, policies seem insufficient to sustain increases in growth of income per capita across the region.
Background: What We Know
Imagine that the position of a country in a cycling race depends on its income per capita. Australia, Hong Kong Special Administrative Region, the United States, and several European economies lead the race (Figure 6.1); some, such as the United States or Canada, have led it for a long time, while others, like Australia, joined the frontline only in recent decades. Another bunch of developed countries (based on income per capita) trails just behind, but is fast catching up—the Asian tigers such as Korea, Taiwan Province of China, and Singapore, alongside New Zealand and Slovenia—all of them growing faster than their predecessors.
Country Ranking on GDP per Capita, 2011
(Purchasing power parity dollars)
Source: IMF, World Economic Outlook database.In a third group, about half as wealthy as the previous group, are a few Latin American countries. Argentina and Chile are leading this group of less-developed economies—together with East European countries—most of them heavily favored by high natural resource prices in the 10 years through 2011.
In proposing guidelines for reaching the lead position, it is useful to compare a handful of leading economies that have abundant natural resources, small domestic markets, and are far from international markets with similar Latin American countries. Before embracing a new development strategy, we argue that the focus should not be on what these leading and similar countries are doing now, but on the strategies they implemented when they had income levels comparable to those in the third group.
The first lesson this analysis reveals is the type of “fuel” the leading economies used to boost growth rates: inspiration, meaning TFP improvements. Latin American economies, by contrast, relied on perspiration to enhance growth rates. During the last decade, saving and investment figures have increased, complemented by lower, if not the lowest, unemployment rates. Yet even though this perspiration approach helped double per capita incomes since 1960, most of the benefits have already been reaped. Comparing economic performance with the United States during 1960–2010, Latin American countries accumulated human and physical capital (12 percent and 16 percent gain, respectively, relative to the United States), but the GDP per capita gap relative to the United States (8 percent loss) has not closed. Again, it is inspiration (TFP growth) that is lacking (29 percent loss).
Moreover, even in countries with sound institutional settings that introduced major structural reforms during the 1980s and 1990s—especially by opening up economies to foreign competition and deregulating financial markets—TFP figures nonetheless suggest that the latter efforts are now showing diminishing returns.2 For Chile, Fuentes, Gredig, and Larraín (2007) show that from 1986 until 1996 steady TFP growth prevailed (together with sound investment and employment rates). But as of 2000, it has stagnated, reaching near-zero average growth. The long duration of depressed TFP growth rates cannot be associated with a particular shock at a point in time. Slower TFP growth explains Chile’s slower output growth, which fell from about 6 percent during the 1980s–1990s to 4.5 percent now. Argentina, Colombia, and Peru show similar patterns.
As Figure 6.2 shows, the TFP contribution to GDP growth is significantly lower in Latin American countries than in industrialized and East Asian countries during 1960–2010, a clear illustration of the perspiration strategy. Public policy should focus on improving TFP, and follow an inspiration strategy to increase long-term growth of income per capita.
Sources of Real GDP Growth, 1960–2010
(Percent)
Sources: Bosworth and Collins (2003); and author’s calculations.Note: TFP = Total Factor Productivity.Recent attention to stronger TFP growth has focused on export structure, especially for those countries with a highly concentrated export bundle, with exports representing the main source of recent growth improvements. While Korea’s export matrix has been sophisticated in the 27 years through 2010, this has not been the case for any Latin American country. Today, chemicals, health-related products, and electronics represent a larger chunk of Korea’s exports, while the share of garments, textiles, and fabrics has declined. The latter products have less value added, because they are less sophisticated in terms of knowledge than the former (Pavitt 1984; Hidalgo and Hausmann 2011). Analyzing a typical Latin American country, two fundamental characteristics emerge. The export pattern is, mostly, heavily biased toward less sophisticated products, such as commodities, agricultural products, and garments, and this pattern has not changed for nearly three decades through 2010.
Countries that have evolved from natural resource exploitation to more sophisticated and heterogeneous export patterns show better growth improvements, in particular in TFP. Thus the Latin American economies face the challenge of adding complexity to their production and export structure, implying less reliance on a few natural-resource-intensive products. Consistent with the previous evidence and based on the production complexity index of Hausmann and others (2011), the Latin American region is not on the frontline of the development race, which is currently led by the Asian tigers and Organisation for Economic Co-operation and Development countries. For instance, Colombia’s production structure is heavily concentrated in a few products related to oil, while Chile’s structure is focused on copper mining. The same pattern holds true for Peru.
Low diversification in product and export structure is related to the slow growth of TFP. According to the World Economic Forum (2013), Chile, for example, ranks quite well on most factors that compose the category of basic requirements for development, such as institutions, infrastructure, and macroeconomic environment. But it underperforms in those pillars related to innovation and business sophistication, which are key for growth in innovation-driven economies. Chile and the Latin American region in general are clearly lagging behind in terms of TFP growth.
In fact, leaving aside the “mean or average approach” and zooming in on the performance of the productive structure within a country, sector heterogeneities pop up swiftly. One can observe highly efficient and export-oriented sectors coexisting with unproductive ones, as can be inferred from the productivity statistics for Chile. Even when zooming in on one particular sector, one can observe different productivity levels among firms. Evidence suggests that this productivity heterogeneity could be represented by a hump-shaped distribution.
Firm and sector heterogeneities have important policy implications. A one-size-fits-all approach will not work when sectors and firms within the same region face dissimilar challenges. Some firms within a sector may be faced with financial constraints, especially the younger and smaller ones, while others may have problems with recruiting qualified human capital. This implies that when designing a new development strategy based on boosting firms’ productivity, heterogeneities need to be taken into consideration. Nevertheless, given market failures, governments should always provide a sound basic institutional setting, which benefits all sectors, such as infrastructure, an educated labor force, and efficient regulation of markets.
Production Diversification and How Innovation May Help
One of the main challenges the Latin American region faces in reaching the frontier of the development race is to reshape its productive structure, taking into consideration the highly heterogeneous performance across sectors and firms within each country (Crespi, Stein, and Fernandez-Aria 2014). The experience of the countries that started from a similar situation in income per capita, natural resource endowment, small domestic markets, and large distance from international markets, shows that more of the same is not enough. Inspiration-led growth could be part of the answer. None of them, for instance Australia, Finland, and New Zealand, pushed a development strategy based on concentrating efforts and productive structure in only a few natural resource products. On the contrary, they explicitly aimed for more diversified economies, increasing local value added in products and services as well as reducing the productivity heterogeneity in local firms.
Four specific country characteristics need to be taken into consideration in the design of a new development strategy:
Size of the domestic markets—One cannot compare a country like Brazil to a country like Peru, because the number of potential customers can make a huge difference for firm and sector development; it is a matter of scale.
Natural resource abundance—This is a fact that the region needs to internalize and to learn how to use wisely. There is clear evidence in some countries of the region showing that the adoption of an industrialization development strategy (and neglecting the comparative advantage of natural resource abundance) may not succeed, as the experience during the 1950s and the 1960s shows (import-substitution strategies). Most Latin American economies are nowadays attempting to build development strategies around their natural resources.
Distance to international markets—This implies a double challenge because transportation and logistics costs go up as distance increases, requiring extra efficiency in certain areas to ensure competitiveness. All sectors face this challenge, but it can get particularly critical for those that export fresh fruits and vegetables, for example.
An enabling institutional setting—This provides the environmental conditions for the productive sector and the society in general to flourish.
However, governments need to be aware that there is no silver-bullet intervention. To tackle these challenges, policymakers need to cope with the paramount task of building an all-encompassing development strategy.
Such a strategy requires the design of a robust conceptual framework that guides the direction of interventions, but always with the vision set on medium- and long-term targets. This framework should take into account the existence of market failures, mainly to ensure that the incentives of the private sector are properly aligned with the overall development strategy. Finally, the strategy should be able to properly balance vertical interventions (specific sectors) and horizontal ones (all sectors), avoiding the risk of government failures. These represent the damage the state can do to markets when trying to solve a market failure.
The introduction of new products and processes is a crucial mechanism to help countries diversify their production structures. It is not only the technical feasibility normally related with invention, but also a solution that creates value. If the production matrix needs to be expanded, by definition, new products should be produced either if they already exist in foreign markets or, as in the case of radical innovations, they are new to the world.
Innovation depends heavily on knowledge, and may need state intervention. Ideas, prototypes, pilot plants, scaling processes, beta versions, packing, patenting, technology transfers, among other activities, are knowledge intensive. This includes scientific, technological, productive, and marketing knowledge, which could be generated through either formal education or by experience (learning by doing). Moreover, knowledge has some economic properties, nonrivalry and nonexclusion, that classify it as a quasi-public good. As such, appropriation problems, asymmetric information, coordination, and other market failures could be raising obstacles when developing and applying knowledge in creating new products. Then, market forces may be insufficient to ensure critical mass for innovation to flourish. Public participation may help create the conditions for this to happen. Moreover, some argue that without state intervention, none of today’s innovations would exist, or at least could have taken considerable time to arise (Mazzucato 2013).
But state interventions have their pitfalls. For a better characterization of state interventions in promoting innovation and production diversity, Crespi and others (2014) present a policy intervention taxonomy as a double-entry matrix (Table 6.1). The vertical axis captures the type of policy and the horizontal axis the scope of policy. The type may consist of the public provision of inputs aimed at solving specific market failures, for instance, setting up some rules of the game. Or policy may consist of a direct intervention in a market to influence the behavior of agents through the design of specific incentives. Scope is related to the pervasiveness of a policy within an economy, that is, if it affects all sectors (horizontal) or specific ones (vertical).
| Policy Scope | |||
|---|---|---|---|
| Horizontal | Vertical | ||
| Type of Policy | Public Inputs | Property rights protection | Phytosanitary controls |
| Market Interventions | R&D tax credits | Tourism tax exemptions | |
| Policy Scope | |||
|---|---|---|---|
| Horizontal | Vertical | ||
| Type of Policy | Public Inputs | Property rights protection | Phytosanitary controls |
| Market Interventions | R&D tax credits | Tourism tax exemptions | |
Different policy instruments are associated with each of the four quadrants of the policy intervention taxonomy. For instance, public inputs providing protection of property rights, particularly relevant for research and development (R&D) and innovation activities, affect all sectors of the economy and thus are considered a horizontal instrument. But specific laws affect particular sectors of the economy; phytosanitary controls, for example, are relevant for agriculture, but not for mining. These input policies are tailor-made for a specific sector and constitute a policy tool to tackle sector heterogeneities. The bottom-left quadrant of Table 6.1 shows policies that directly intervene in the markets with the aim of modifying the incentives of agents to influence their behavior. They can be horizontal or vertical. Under the former, for example, R&D tax credits change the cost that firms face when engaging in R&D, naturally modifying their assessment of expected returns on R&D projects. Other interventions may be aimed at specific sectors, such as tourism.
Vertical intervention policy tools have to be used carefully to avoid generating government failures. And under some circumstances, it is preferable to leave a market failure unattended than to intervene, because the costs may exceed the benefits.
Given the risks associated with policies of vertical scope, countries have preferred horizontal approaches to policymaking, focusing on enabling an institutional setting (rules of the game) that benefits all sectors of the economy. Once this is done, however, the specificities of each sector need to be dealt with, and then a vertical approach is required.
When doing so, policymakers need to avoid at least three types of state failure:
Time inconsistency—This relates to policymakers’ tendency to favor policies more visible to the citizenry in the short term, like building a bridge, because it may increase reelection possibilities. Because science, technology, and innovation policies, required to promote diversification of the production structure, are quite invisible compared with, say, a bridge, and yield benefits in the medium or long term, they are less likely to be implemented by policymakers focused on their political parties’ reelection. This may be especially relevant within the short-term nature of the government settings. To avoid this, science, technology, and innovation and other production promotion policies should be implemented by state organizations with specific mandates to ensure time consistency.
Agency problems. This occurs when the interest of the principal, say a president or prime minister, differs from the interest of a minister, such as the minister of industry. The problem is that the objectives of the principal and the agent may differ, introducing distortions between overall planning and implementation. Usually the agent that is closer to the implementation scene has more practical information about the beneficiary of a given policy or public instrument. Another derivative of the agency problem has to do with the absence of a clear responsibility for the policy implementation, which occurs given the interdisciplinary nature of the policy and tasks related to different ministries. Not having a clear definition of roles in science, technology, and innovation and production diversification policy intensifies the potential for dynamic inconsistency problems.
Public capture. This occurs when the state is captured by the beneficiaries, who end up influencing the course and direction of some policies. A common case is the implementation of an intervention initially justified by a market failure. This intervention may take the shape of a subsidy aimed at a specific group. Once the failure has been resolved, however, the beneficiaries may lobby in favor of the continuation of the measure, in exchange for votes, for example.
Because of the threat of government failures, policies require a proper public institutional setting that minimizes their manifestation. This is what Latin American countries have been trying to do in the decade through 2014. The approach undertaken has consisted of dividing the strategy setting, policy design, and implementation tasks among different bodies of the public apparatus. The institutional setup for policy should be ideally composed of, at the first level, an advisory body in charge of defining the medium- and long-term objectives of policies, perhaps a chief executive officer of innovation or a state strategy agency properly insulated from political cycles. At the second level, a ministry, or a few ministries, would be in charge of policy design, based on the guidelines given by the advisory body; and finally, the third level would involve a public agency in charge of the direct implementation of the science, technology, and innovation and production diversification policy instruments.
What Science, Technology, and Innovation Can Do
Latin American countries should learn three lessons about innovation. First, science and technology is not just a rich-country hobby. A large body of theoretical (Romer 1990; Aghion and Howitt 1992) and empirical (Grilliches 1995; Hall and Jones 1999; Rouvinen 2002) literature suggests that countries that have invested in R&D and related activities have achieved higher TFP growth rates and subsequent increases in income per capita.
The second lesson is that the effects of R&D efforts over TFP enhancement have a natural delay. That is, one may observe a drop in TFP in the short term, but it would eventually go up in the medium and long term (Goto and Suzuki 1989; Benavente, de Gregorio, and Nuñez 2006). This delay is associated with a natural learning process that goes hand in hand with the innovation process, as Schumpeter emphasized when he developed his idea of “creative destruction.” This idea is quite intuitive for engineers who are aware that there is a natural adaptation process when a new machine is incorporated into the production process. But economists don’t seem to have understood this very well. In fact, the data show that increases in TFP take about two years in the manufacturing sector and eight years in the biotechnology sector. Heterogeneity is unveiled in yet another dimension.
The third lesson is that competition problems are more prone to occur in a small market setting. Since competition is a necessary condition to promote innovation, smaller economies need to deal with this intrinsic limitation. A common strategy of smaller countries currently at the technological frontier has been to compete in global markets, not in domestic ones. The only way to keep up with fierce competition in international markets is to invest in R&D and innovation, which at the aggregate level will ultimately boost productivity growth (Bravo-Ortega, Benavente, and Gonzalez 2014). The latter implies a different innovation-TFP causality direction for small and large economies. The large ones, like Brazil or the United States, start increasing TFP and then become competitive within local markets, after which they begin exporting.
It is important to mention the side effects of the innovation process. Contrary to the common expectation that the innovation will decrease employment, evidence suggests that this is not the case (Harrison and others 2014; Mohnen and Hall 2013; Benavente and Lauterbach 2008). The introduction of innovations that involve incorporating machinery that may eventually substitute human capital for physical capital may affect employment. However, a more efficient firm may face higher demand for their product, triggering an increase in the demand for labor—and probably more qualified labor. Eventually, we observe a net increase in labor demand derived from the innovation process, especially for less-developed economies.
Finally, the innovation process may have important implications for income distribution issues so prevalent in Latin American economies. Because innovation goes hand in hand with learning, knowledge becomes of the utmost importance for these processes to occur. Because knowledge has public good characteristics, its rents tend to be dispersed more widely, as opposed to rents derived from land or physical capital. This has important implications when designing strategies to address income inequalities.
Conclusion
Recent events have shown that most Latin American countries remain vulnerable to the ups and downs of the global economy, especially commodity price fluctuations. Countries that elsewhere have succeeded in leapfrogging the middle-income trap have pushed inspiration-based policies. Those that put the different manifestations of knowledge at the center of their public policy goals have done well. They have focused on productivity enhancements, mission-oriented sciences, innovation, and entrepreneurship, among others. These countries have also introduced explicit mechanisms to diversify their production matrix, mostly based on the introduction of new products and more efficient processes.
Product diversification is not a goal in itself, but a means to leap into the frontline of the development race. The Latin American experience also tells us no single, silver-bullet policy can achieve this. On the contrary, a systemic and dynamic approach to policymaking needs to be adopted to successfully implement a development strategy that enables these economies to achieve their goals in the medium and long term.
In particular, the institutional setting under which the new strategies for development are implemented becomes of the utmost importance. Given that state failures can threaten the success of a well-designed development strategy, countries need to make an effort to design and establish an adequate institutional setting, no matter how difficult it is.
Policymakers, and citizens generally, must not forget that the results and benefits of science, technology, and product diversification policies may only become tangible in the medium and long term. Countries need to design institutions, explicitly and implicitly, that can help to avoid this anxiety, a characteristic that also explains the difference between developed economies and those that are lagging behind.
References
Aghion, P., and P. Howitt. 1992. “A Model of Growth through Creative Destruction.” Econometrica 60 (2): 323–51.
Benavente, J. M., J. de Gregorio, and M. Núñez. 2006. “Rates of Return for Industrial R&D in Chile.” Working Paper No. 220, Department of Economics, University of Chile.
Benavente, J. M., and R. Lauterbach. 2008. “Technological Innovation and Employment: Complements or Substitutes?” European Journal of Development Research 20 (2): 318–29.
Bosworth, B., and S. Collins. 2003. “The Empirics of Growth: An Update.” Brookings Papers on Economic Activity 34 (2): 113–206.
Bravo-Ortega, C., J. M. Benavente, and A. Gonzalez. 2014. “Innovation, Exports and Productivity: Learning and Self Selection in Chile.” Emerging Markets Finance and Trade. 50 (1): 68–95.
Crespi, G. 2006. “Productivity and Firm Heterogeneity in Chile.” PRUS Working Paper 36, Poverty Research Unit, University of Sussex.
Crespi, G., E. Stein, and E. Fernandez-Arias, eds. 2014. Rethinking Productive Development: Sound Policies and Institutions for Economic Transformation. London: Palgrave Macmillan.
Fuentes, R., F. Gredig, and M. Larraín. 2007. “Estimating the Output Gap for Chile.” Working Papers Central Bank of Chile 455, Central Bank of Chile.
Goto, A., and K. Suzuki. 1989. “R&D Capital Rate of Return on R&D Investment and Spillover of R&D in Japanese Manufacturing Industries.” Review of Economics and Statistics. 71 (4): 555–64.
Grilliches, Z. 1995. “R&D and Productivity.” In Handbook of the Economics of Innovation and Technological Change, edited by P. Stoneman, 52–89. Oxford: Blackwell.
Hall, Robert E., and Charles I. Jones. 1999. “Why Do Some Countries Produce So Much More Output per Worker Than Others?” Quarterly Journal of Economics. 114 (1): 83–116.
Harrison, R., J. Jaumandreu, J. Mairesse, and B. Peters. 2014. “Does Innovation Stimulate Employment? A Firm-Level Analysis Using Comparable Micro-Data from Four European Countries.” International Journal of Industrial Organization 35 (C): 29–43.
Hausmann, R., C. A. Hidalgo, S. Bustos, M. Coscia, A. Simoes, and M. Yildirim. 2011. The Atlas of Economic Complexity. Hollis, New Hampshire: Puritan Press.
Hidalgo, César A., and Ricardo Hausmann. 2011. “The Network Structure of Economic Output.” Journal of Economic Growth 16: 309–42.
Mazzucato, M. 2013. The Entrepreneurial State: Debunking Private vs. Public Sector Myths. London: Anthem Press.
Mohnen, P., and B. Hall. 2013. “Innovation and Productivity: An Update.” Eurasian Business Review. 3 (1): 47–65.
Pavcnick, N. 2002. “Trade Liberalization, Exit and Productivity Improvements: Evidence from Chilean Plants.” Review of Economic Studies 69 (1): 245–76.
Pavitt, K. 1984. “Sectoral Patterns of Technical Change: Towards a Taxonomy and a Theory.” Research Policy 13 (6): 343–73.
Roberts, M. J., and J. R. Tybout, eds. 1996. Industrial Evolution in Developing Countries: Micro Patterns of Turnover, Productivity and Market Structure. New York: Oxford University Press.
Romer, P. 1990. “Endogenous Technological Change.” Journal of Political Economy 98 (5): 71–102.
Rouvinen, P. 2002. “R&D-Productivity Dynamics: Causality, Lags, and ‘Dry Holes’.” Journal of Applied Economics 5 (1): 123–56.
World Economic Forum. 2013. Global Competitiveness Report, 2012–2013. Geneva: World Economic Forum.
Essentially due to skills mismatch between workers and firms.
As shown by Pavcnick (2002) and Crespi (2006) in the case of Chile, and Roberts and Tybout (1996, 227–58) for Colombia.
