Development and Trade. The Role of Learning, Investment and Technology

Content

1. Introduction

2.1 Model and empirics on development, investment and learning
2.1.1 1st Model – theoretical part
2.1.2 1st Evidence – empirical part
2.2 Model and empirics on development, specialisation and trade
2.2.1 2nd Model – theoretical part
2.2.2 2nd Evidence – empirical part

3. Conclusion

References:

1. Introduction

In order to achieve economic development, here understood as GDP growth, it is often suggested that developing countries would only need to ensure free trade and good institutions. Good institutions shall guarantee for efficient technology adaption and enact property right laws. Free trade shall enable countries to fully enjoy the gains from trade. Nevertheless countries pursuing institutional quality and trade openness, namely in Latin America, have failed to achieve sustainable economic development. Meanwhile others, namely in South and East Asia, have achieved remarkable economic growth rates, without functioning markets and property rights in place (Hausmann/Rodrik 2003, p.607).

Recent literature on development economics tries to explain that diverging performance. It picks up the simple fact that developing countries, more precisely their entrepreneurs, first and foremost need to find out what sectors, goods and technologies to specialise in. This inherently bears the question of in how many sectors, goods and technologies to engage, so the optimal degree of diversification.

In the 1st paper, Dani Rodrik and Ricardo Hausmann try to shed light onto that issue. They focus on how to stimulate this discovery-process. Their work is based on some assumptions. After the discovery process entrepreneurs may learn from each other. The private rent to a discovery is smaller than the social rent. That is why policy needs to internalise some of the rent back to the investor. This can potentially be achieved by hampering imitated entry into the newly discovered market. Nevertheless governments in developing countries need to act in a balanced way, as that exclusivity always comes along with social costs. Rodrik and Hausmann consider different policy options in order to address those problems.

The questions I try to answer are: How can a theoretical framework be designed to explain the mentioned inefficiencies of the laisez-faire equilibrium in comparison to a socially optimal situation? Can it be argued that the above mentioned diverging growth success of different developing countries has been due to varying quality of chosen policy options? Do, by their enhancement, Rodrik and Hausmann explain the seemingly anomalies within the established theory on development economics?

With respect to the 2nd paper, Dani Rodrik, Ricardo Hausmann, and Jason Hwang stay in the framework assuming positive externalities in a cost-discovery process in the sense of domestic imitation and mutual learning of entrepreneurs. The model is concerned with the effects of an economy’s maximal discovered productivity on its economic growth rate. Further, the determinants of this expected maximal productivity are of interest. The authors’ in particular base their argument on production indeterminacy. Though in a Heckscher-Ohlin model with more goods than factors endowment influences the specialisation pattern, it does not pin it down uniquely (Dixit/Norman 1994, p. 119). This leaves a potentially positive role for governmental economic policy to play. Assuming economic growth to be a desirable aim and industrial policy to be a legitimate instrument, it opens the door to policy issues, namely:

Is it possible to foster economic growth by a clever choice of specialisation pattern? Can an economy achieve growth by specialising on high productivity goods? And to sum up: Can a developing country achieve economic growth by specialising on a certain kind of goods?

Conceptually and with regard to empirical evidence a challenge of the 2nd paper is, whether there are types of goods which are usually produced by rich countries and what those goods’ characteristics in terms of productivity are.

I present the two models and the respective evidence in a theoretical and an empirical part for each model. The latter part of the analysis focuses on a critical examination of the empirical evidence provided on the proposed causal relation between governmental policy, specialisation pattern and economic growth. In the conclusion I summarize the results and identify the issues which are still open.

2.1 Model and empirics on development, investment and learning

2.1.1 1st Model – theoretical part

In the 1st paper, as key elements of their argument Rodrik and Hausmann identify uncertainty about production pattern, difficulties in importing technology and rapidness of domestic imitation (Hausmann/Rodrik 2003, p. 614).

Uncertainty about production pattern refers to the theroretical predictions in Heckscher-Ohlin models, that a country’s factor endowment influences but not uniquely determines its specialisation pattern. If there are more output goods than input factors, the full employment conditions for the number of factors do not exactly solve for the output quantities (Feenstra 2004, p. 84). This opens the door to use the remaining space of production possibilities to specialise in a way, on those sectors, goods and technologies that highest possible GDP growth rates are achieved. The paper by Rodrik and Hausmann is based on implicitly equalising development with GDP growth.

Difficulties in importing technology picks up the observation that foreign technology, although it is in place abroad, cannot be instantly put to work domestically (Rodrik 2007, p. 163). Importing technology requieres domestic tinkering, which is associated with costs. Foreign technology needs to be adapted to local conditions in order to work efficiently. The impossiblity of importing technology off-the-shelf rules out free trade as sufficient strategy for economic growth. Other policy options must be considered.

Rapidness of domestic imitation is a basis of Rodrik and Hausmann’s argumentation, as positive externalities of investment are key (Hausmann/Rodrik 2003, p.605). The faster domestic imitation occurs, the lower is the private rent of the initial entrepreneur, who entered the newly discovered market first. Fast domestic imitation means low private returns with unchanged higher social returns. In order to stimulate the discovery process private returns of innovators must be increased. Several policy options are possible.

Rodrik and Hausmann run their model for three different scenaria, two of which are centralised social planner problems, the remaining third case describes a decentralised market solution. The social planner scenaria capture full and incomplete information cases, though the incomplete information situation is arguably the more realistic one. A comparison across scenaria with respect to the amount of investment incurred in the 1st period and the number of different goods produced in the 3rd period will allow for evaluation.

The model assumes a small open economy with two sectors, a modern and a traditional one. The distinction criteria between sectors is whether costs of production are known.

The modern sector is characterised by an unknown productivity. The productivity parameter [Abbildung in dieser Leseprobe nicht enthalten] is uniformly distributed over the intervall [Abbildung in dieser Leseprobe nicht enthalten] and independant and identically distributed over the [Abbildung in dieser Leseprobe nicht enthalten] potential modern goods. It is property of the individual good [Abbildung in dieser Leseprobe nicht enthalten]. The unit costs of production increase with the wage rate and the number of workers needed to produce a single unit of a particular good, [Abbildung in dieser Leseprobe nicht enthalten]. An economy-wide wage rate [Abbildung in dieser Leseprobe nicht enthalten] implies intersectoral mobility of labour. Uncertainty over the productivity parameter of good [Abbildung in dieser Leseprobe nicht enthalten] captures first, learning about how to combine different inputs, second, figuring out whether production can be incurred efficiently, and third, exploring what the costs of production are. The modern sector’s production function uses labour as only input factor and operates under constant returns to scale. The [Abbildung in dieser Leseprobe nicht enthalten] entrepreneurs who chose to invest in the cost discovery of the modern sector run exactly one firm each. By normalisation, the size of each firm equals one unit of output of good [Abbildung in dieser Leseprobe nicht enthalten].

Three periods are distinguished. The 1st period marks the discovery process. Each of the [Abbildung in dieser Leseprobe nicht enthalten] entrepreneurs invests in the productivity discovery of one of the [Abbildung in dieser Leseprobe nicht enthalten] potential modern sector goods. The 2nd period stands for production at monopoly profits. Entrepreneurs whose good can be produced below exogenous world market price [Abbildung in dieser Leseprobe nicht enthalten]earn excess profits, entrepreneurs whose good cannot be profitably produced close their firm without extra costs. The period of monopoly profits can be motivated first, because it may take time that productivity parameters become common knowledge (delayed learning), and second, because it may take time that emulators set up a firm (delayed imitation). The final 3rd period entails production with free entry at zero profits. New entrepreneurs can enter the modern sector, learn and imitate the modern goods. Profits are eliminated via an upward adjustment of wages.

The traditional sector’s production function is characterised by constant returns to scale, where labour and a fixed factor are used as inputs, [Abbildung in dieser Leseprobe nicht enthalten]. [Abbildung in dieser Leseprobe nicht enthalten] is the economy’s total labour force, [Abbildung in dieser Leseprobe nicht enthalten] employment in the modern sector, and [Abbildung in dieser Leseprobe nicht enthalten]the factor share of labour in the traditional sector. [Abbildung in dieser Leseprobe nicht enthalten] is the value marginal product of labour in the traditional sector, which is positive but diminishing in the traditional sector’s labour employment. A single homogenous traditional good is produced whose price is assumed to be [Abbildung in dieser Leseprobe nicht enthalten].

The 1st scenario captures social optimality in the full-information case. The optimising agent is a social planner that enjoys full knowledge about goods’ productivity parameters. Once productivities have been discovered, the social planner produces no other modern good except the one with maximal productivity. The optimality conditions that are entailed in this problem capture the optimal allocation of labour across sectors. Useful for that purpose is a comparision of the value marginal product of labour in the traditional sector with the value marginal product of labour in the modern sector. The marginal product of labour in the traditional sector is the 1st derivative of the production function with respect to labour. It is maximal if all labour of the economy is employed in the traditional sector, and it is minimal if the traditional sector does not employ any labour at all. In the modern sector, the value marginal product of labour is given by the modern good’s price times the units of the most productive good’s output per investor’s labour input.

Abbildung in dieser Leseprobe nicht enthalten

If the value marginal product of labour for the most productive good in the modern sector lies within the range of (value) marginal products of labour in the traditional sector, then labour is partly employed in the modern sector. There is diversified production of the most productive good in the modern sector and the traditional sector’s good.

If the value marginal product of labour for the most productive good in the modern sector does not lie within the range of (value) marginal products of labour in the traditional sector, then there is full specialisation either on the traditional sector or the modern sector’s most productive good.

The 2nd scenario captures social optimality in the incomplete information case. Optimising agent is a social planner who has to incurr same learning costs with respect to discovering productivity as private agents. The planner is assumed to be risk neutral. Once productivities have been discovered, the social planner produces no other modern good except the one with maximal productivity. The expected maximum productivity increases at diminishing rate with the number of entrepreneurs, [Abbildung in dieser Leseprobe nicht enthalten].

During the 1st period production only takes place in the traditional sector, while there is cost discovery in the modern sector. The social planner allows free entry already in the 2nd period, so there is no difference between 2nd and 3rd period results. Diversified production in traditional sector and modern sector’s most productive good is assumed.

Traditional sector’s output in the 1st period is [Abbildung in dieser Leseprobe nicht enthalten] and decreases with the number of investments. Modern sector’s 1st period employment [Abbildung in dieser Leseprobe nicht enthalten] is the number of investments times employment per investment.

Modern sector’s expected output in the 2nd period is [Abbildung in dieser Leseprobe nicht enthalten], traditional sector’s expected output in the 2nd period is [Abbildung in dieser Leseprobe nicht enthalten]. Modern sector’s 2nd period employment [Abbildung in dieser Leseprobe nicht enthalten] derives from equalisation of (expected) value marginal products of labour across sectors. Equalisation of (expected) value marginal products across sectors reflects optimal behaviour of the social planner. The (value) marginal product of labour in the traditional sector is given by the first derivative of the traditional production function with respect to labour. In the modern sector, the expected value marginal product of labour is given by the modern good’s price times the expected units of the maximal productive good’s output per investor’s labour input.

Abbildung in dieser Leseprobe nicht enthalten

The social planner maximises total wealth [Abbildung in dieser Leseprobe nicht enthalten] over the optimal choice of investment in cost discovery in the 1st period. Total wealth depends on (expected) output in modern [Abbildung in dieser Leseprobe nicht enthalten] and traditional [Abbildung in dieser Leseprobe nicht enthalten] sector, where the latter is active in both periods. [Abbildung in dieser Leseprobe nicht enthalten] is the discount factor between 1st and 2nd period and [Abbildung in dieser Leseprobe nicht enthalten] the instantaneous discount rate .

[...]