Consumer Behavior and Public Policies: Empirical Evidence through vec model on Brazil’s Automotive Industry

Consumer Behavior and Public Policies:

Empirical Evidence through VEC model on Brazil’s Automotive Industry

Gérson Guilherme Lima Linhares¹

Eveline Barbosa Silva Carvalho²

Abstract
The study analyzes the effects on consumer behavior of public policies using empirical evidence on Brazil’s automotive industry. Policies adopted by the Brazilian government to encourage the acquisition of vehicles through tax reduction on industrialized products and bank credit were estimated using an error correction vector model. Simulations on sales change were performed and results showed relevant increase in two different periods. The model showed that the policies caused an increase of 23,2% and 16,8% on sales on the first and second periods analyzed with significant impacts to the market. The Tax on Industrialized Products reduction showed more convincing to make consumers buy vehicles than bank credit. Results lead to the conclusion that the direct impact on demand increase and the artificial incentive to producers may induce wrong management decisions in the long run.
Keywords: Tax on Industrialized Products, Automotive Industry, Error Correction Vector Model.

JEL Classification : D03; L62; C32.

Resumo

O estudo analisa os efeitos políticas públicas sobre o comportamento do consumidor, por meio do uso de evidência empírica sobre a indústria automotiva do Brasil. As políticas adotadas pelo governo brasileiro para incentivar a aquisição de veículos por meio da redução do imposto sobre produtos industrializados e de crédito bancário foram estimadas usando um modelo de vetor de correção de erros. As simulações sobre mudança vendas foram realizadas e os resultados mostraram aumento relevante em dois períodos diferentes. O modelo mostrou que as políticas causaram um aumento de 23,2% e 16,8% nas vendas sobre,respectivamente, o primeiro e segundo períodos analisados com impactos significativos para o mercado. A redução do Imposto sobre Produtos Industrializados mostrou mais convincente para fazer os consumidores compram veículos de crédito bancário. Os resultados levam à conclusão de que o impacto direto no aumento da demanda e o incentivo artificial aos produtores pode induzir decisões de gestão erradas no longo prazo.

Public Policies are common around the world and have been present in Brazil for many years. There are many kinds of policies and governments generally apply them to benefit producers in response to pressure from interest groups (Becker, 1983).

The automotive industry is one of those groups that have been in Brazil for more than 50 years and represent such an important market that the share of industrial Gross Domestic Product (GDP) represents around 19% and the number of employees was over 150 thousand in 2012 according to ANFAVEA (2014).

Arguing recession and difficulties to pay salaries and maintain employees this industry lobby the government that as a response apply protectionist policies. In fact, in recent years, government intervened several times in the auto industry.

One form of intervention has been to reduce the tax rates on industrialized products (IPI) for locally manufactured vehicles, in order to stimulate sales internally and avoid the drop in the number of workers employed. Another form of intervention has been through the reduction of the tax on financial transactions (IOF) in loans to stimulate credit by banks to individuals wishing to purchase vehicles.

This study aims to analyze the impact of public policies using as a study case the reduction of the IPI and IOF on the sales of cars and light commercial vehicles produced in Brazil in two periods: between January 2009 and March 2010, which was the period of the first IPI reduction and between June 2012 and December 2012, which was the period of the second IPI reduction. We seek to find out the impacts of these two policies on the sales increase and to what extent this market intervention may impact the sector in the long run.

To achieve the objectives an Error Correction Vector Model (VECM) will be estimated, where the time series on sales, prices, credit and income are in logarithmic form, to facilitate the study of the elasticity of prices, credit and income to sales, and also to obtain long-term relationship between the variables. After that, simulations of sales behavior considering three scenarios will be made for the two referred periods. The first scenario, do not consider IPI reduction, which will be useful to compare the impacts with reduced IPI, the second scenario, considers IPI reduction together with 5% increase in credit grant, which will provide the impact of credit with reduced IPI and the third scenario do not consider IPI reduction but only a 5% increase in credit grant, which will show the impact of the credit without IPI reduction.

One of the justifications for testing the vehicles industry is the fact that despite being such an important market there are few empirical studies on the impacts policies on sales of vehicles.

The study aims to contribute to a better understanding of the quantitative impact of public policies, such as the IPI reduction, adopted by the Brazil´s federal government in recent years in order to boost domestic consumption. However the study does not focus on the consequences of policies in terms of tax revenues or how much it impacts the government in terms of budget. The relevance of this study is to contribute to understanding the consequences in terms of sales which affects consumer decision and company management in the long run.

Following this introduction, the paper is divided in six sections. Section two is devoted to the understanding of the study case to be tested, which includes the trajectory of Brazilian automotive industry and policies adopted by the Brazilian government on the sector. Section three presents a literature review on the theoretical framework and empirical studies. In section four, there is a description of methodology including the data with respective sources, as well as methodological procedures adopted to use a Vector Error Correction Model (VECM). Section five presents the model results and Simulations of vehicle sales and at last the final considerations.

Between 2004 and 2007 Brazil had an average annual growth of 13% in production and sale of vehicles. However, on September 2008, while the total sales of motor vehicles exceeded in about 30% the sales of the same month in the previous year, the production grew by approximately 20%. (Alvarenga et al., 2010b, p.8).

According to Barros and Pedro (2011), the increase in sales was due to the average income growth of Brazilians, the upward mobility of part of the population to class C, which helped many of them buy the first car, reduction on unemployment levels and better credit access with lower interest rates and longer financing terms.

With the 2008 crisis, there was a sharp reduction in the production and sales of vehicles. In December 2008 vehicle production was 47.1% lower than in November of that same year (Poguetto, 2009).

Figure 1, shows that the number of cars and light commercial licensees, considered as a sales approach, had a large reduction on the second half of 2008 in comparison to the first half. In fact, while in July 2008, there were about 237,000 licensed vehicles in December of the same year only 153,000 vehicles were licensed, representing a reduction of approximately 35.4% in six months.

According to Alvarenga et al. (2010a, 2010b), companies in the automotive industry reacted to the crisis by reducing work shifts and granting collective holidays to workers.

In the first half of 2009, there was a recovery but the number of licensed cars and light commercial vehicles would only overcome the data of July 2008 in June 2009 when nearly 250,000 vehicles were licensed. And during 2010, despite the decline in the first two months, in March 2010 sales reached almost 280,000 units.

Between January 2009 and March 2010 about 208,000 vehicles were sold on average per month, 10% higher than the monthly average in the second half of 2008 (between July and December), when 189,000 vehicles were sold, on average.

With respect to credit, there was a reduction in the volume of grants for the purchase of vehicles during the second half of 2008. While in July 2008 there was approximately R$4.75 billion in credit contracts, in December of the same year there was only approximately R$2.94 billion, representing a reduction of 38% in just six months. Starting in January 2009, there has been a resumption of credit concessions and in June 2009 the value of credit operations reached approximately R$ 5.39 billion, exceeding the month of July 2008 in almost 13.3%.

Credit concessions between January 2009 and March 2010 had a monthly average of R$ 5.574 billion, 56% higher than the monthly average of the second half of 2008, when the monthly average was only R$ 3.561 billion.

During January and May 2012 the loans were on average R$ 7.355 billion, while between June and December 2012, credit concessions averaged R$ 8.152 billion, meaning an increase of approximately 10.8% of the volume of concessions, between these two periods of the year 2012.