Annual Report 2018

44 Economics SERVICES, INDUSTRIAL PRODUCTIVITY AND AGGREGATE PRODUCTIVITY IN BRAZIL The service sector is dominant in terms of value added and labor, but typically receives little attention compared to the industrial and agricultural sectors. This study develops a theoretical model in which services are not only a final consumer good, but also an intermediate good, and evaluates the repercussions on the productivity of the other sectors of the economy. OBJECTIVE • To quantitatively evaluate the effect of services, considered to be intermediate goods, on aggregate productivity in Brazil and their impact on the productivity of the other sectors, particularly industry. RESEARCH METHOD • We developed a general equilibrium model in which there are three types of goods and sectors of the economy: agriculture, industry and services. For each sector, firms produce goods destined for consumption of individuals or utilized by other firms as intermediate goods. Firms decide how much labor to employ and the volume of intermediate goods to use. The economy is inhabited by an infinite number of homogeneous agents, who offer labor and decide how much of each good to consume. According to the method normally used in models like this, it is assumed that some pa- rameters are directly observed in the data, while others, which do not have a clear cor- respondence with the data, are estimated based on the model's equilibrium conditions. In the computable general equilibrium model, services are not only consumption goods, they also serve as intermediate goods. The idea is that the firm in the industrial sector has a production function that uses inputs from labor and intermediate goods. In this case, low productivity of services affects the productivity of industry and of the service sector itself. The model was calibrated and simulated for the Brazilian economy,, closely reproduc- ing the data, as for example, aggregate labor productivity and the distribution of labor among the sectors. Counterfactual exercises were implemented in which a parameter or exogenous variable was modified. The objective of all the exercises was to understand the impact of sectoral changes on the economy as a whole, as well as to measure and compare the gains from changes in total factor productivity (TFP) and in the productive structure of the econo- my. Unlike what happens in models without intermediate goods, a multiplicative effect now exists that those models do not capture and that can have sizable effects on the economy and sectoral productivity. of the American economy, while maintaining the sectoral structure constant and equal to that calibrated for the Brazilian economy. This result is in the same direction as results obtained with very different methods. Al- though the values are not the same, the orders of magnitude in the two experiments are close to each other. This confirms, now in a more detailed theoretical model, that the problem of low productivity in Brazil is more a problem of level, with all sectors having weak productivity, than of composition. This means that Brazil is not specialized in sec- tors with low productivity, but that productivity in all sectors of the Brazilian economy is relatively small. CONTRIBUTIONS OF THE STUDY . The experiment is important to establish the relevance of including intermediate goods when one is interested in investigating the impacts of sectoral changes - mainly of varia- tions in productivity of services - on the aggregate productivity of the economy. APPLICATIONS OF THE RESULTS AND POSSIBLE EXTENSIONS OF THE STUDY The project sheds light on the sectoral relations in the Brazilian economy, providing in- struments to attain more effective policies for growth in Brazil. Given that all economies tend to shift more to services as they develop, better knowledge of this sector's func- tioning and its determinants is a necessary condition for implementing effective policies in Brazil. RESULTS • The results show that a change in the productive structure alone would induce a small increase in productivity. With all else constant, if Brazil had coefficients of the production function equal to those of the United States - and consequently elasticities of production in relation to intermediate goods equal to those of that country, productivity would be 21% higher than observed in the data. Brazil's relative productivity would jump from 14.4% to 17.5% of that of the USA, a modest gain. However, although relevant, this gain is significantly lower than what would be observed by attributing to each of the three sectors the productivity of the corresponding sectors AUTHORS: Fernando Veloso, Pedro Cavalcanti Ferreira and Bruno Ricardo Delalibera. ORGANIZATION: Brazilian Institute of Economics (IBRE), Center for Growth and Economic Development Studies (Growth and Development) and EPGE Brazilian School of Economics and Finance (EPGE). Annual Report 2018 45 RESEARCH

View entire presentation