Portal de Periódicos da CAPES

Artigo Produção Nacional Revisado por pares

BIBTEX RIS

Measuring Micro- and Macro-Impacts of Regional Development Policies: The Case of the Northeast Regional Fund (FNE) Industrial Loans in Brazil, 2000–2006

2012; Routledge; Volume: 48; Issue: 4 Linguagem: Inglês

10.1080/00343404.2012.667872

1360-0591

Guilherme Resende,

Fiscal Policy and Economic Growth

Abstract Resende G. M. Measuring micro- and macro-impacts of regional development policies: the case of the Northeast regional fund (FNE) industrial loans in Brazil, 2000–2006, Regional Studies. This paper formulates a framework to measure the micro- and macro-impacts of regional development policies in Brazil using the first-differences method that controls for observable characteristics and unobserved fixed effects. Next, it applies this framework to measure the impact of the Northeast regional fund (FNE) industrial loans on employment and labour productivity growth at the micro- (firm) level and on gross domestic product (GDP) per capita growth at macro- (municipalities, micro-regions and spatial clusters) levels for the 2000–2003 and 2000–2006 periods. The results show a positive and statistically significant impact of the FNE industrial loans on job creation at the micro-level but no significant impacts on the GDP per capita growth at the macro-level. Resende G. M. 区域发展政策宏观和微观影响测度:以巴西东北地区融资产业贷款为例(2000–2006),区域研究。本文通过一阶差分法控制可观测特征和不可观测固定效应,构建了测度巴西区域发展政策宏观及微观影响的分析框架。之后,本文利用这一分析框架测度了2000–2003 年及2003–2006 年两个时段中,巴西东北地区融资产业贷款在微观尺度(企业层面)对就业情况及劳动生产率的影响,以及在宏观尺度(市、小尺度区域和空间集群层面)对人均国内生产总值的影响。结论表明,这一政策在微观尺度对创造就业机会起到了积极的、统计上显著的影响,但在宏观层面对人均国内生产总值的影响并不显著。 Resende G. M. Mesurer les impacts micro- et macro-économiques de la politique d'aménagement du territoire: étude de cas des prêts industriels au Brésil distribués sous l'égide du Fonds régional du Nord-Est (FNE), de l'an 2000 à 2006, Regional Studies. Cet article cherche à élaborer une structure pour mesurer les impacts micro- et macro-économiques de la politique d'aménagement du territoire au Brésil employant la méthode de la différence d'ordre 1 qui tient compte des caractéristiques observées et des effets fixes non-observées. Ensuite, on se sert de cette structure afin de mesurer l'impact des prêts industriels distribués sous l'égide du Fonds régional du Nord-Est (FNE) sur l'emploi et sur la croissance de la productivité du travail au niveau micro-économique (à savoir, les entreprises) et sur la croissance du Produit intérieur brut (PIB) par tête au niveau macro-économqiue (à savoir, les municipalités, les micro-régions et les clusters géographiques) pour les périodes allant de l'an 2000 jusqu'à 2003, et de l'an 2000 à 2006. Les résultats laissent voir une corrélation étroite et statistiquement significative entre les prêts industriels distribués sous l'égide du FNE et la création de l'emploi au niveau micro-économqiue, mais peu de corrélation significative sur la croissance par tête du PIB au niveau macro-économique. Resende G. M. Messung der Auswirkungen von regionalen Entwicklungspolitiken auf Mikro- und Makroebene: der Fall der Industriedarlehen des Regionalfonds Nordost (FNE) in Brasilien, 2000-2006, Regional Studies. In diesem Beitrag wird mit Hilfe der First-Difference-Methode unter Berücksichtigung von beobachtbaren Merkmalen und nicht beobachteten Festeffekten ein Rahmen zur Messung der Auswirkungen von regionalen Entwicklungspolitiken auf Mikro- und Makroebene in Brasilien formuliert. Anschließend wird dieser Rahmen zur Messung der Auswirkung der Industriedarlehen des Regionalfonds Nordost (FNE) auf das Wachstum des Beschäftigungsniveaus und der Produktivität der Arbeitnehmer auf Mikroebene (Firmen) sowie auf das Wachstum des Bruttoinlandsprodukts (BIP) pro Kopf auf Makroebene (Gemeinden, Mikroregionen und Raumcluster) für die Zeiträume von 2000 bis 2003 und von 2000 bis 2006 eingesetzt. Aus den Ergebnissen geht eine positive und statistisch signifikante Auswirkung der Industriedarlehen des FNE auf die Schaffung von Arbeitsplätzen auf Mikroebene hervor, nicht jedoch eine signifikante Auswirkung auf das Wachstum des Pro-Kopf-BIP auf Makroebene. Resende G. M. Medición de las repercusiones de las políticas de desarrollo regional a nivel micro y macro: el caso de los préstamos industriales del Fondo Regional Noreste (FNE) en Brasil, 2000–2006, Regional Studies. En este artículo proponemos un esquema para medir la repercusiones a nivel micro y macro de las políticas de desarrollo regional en Brasil utilizando el método de las primeras diferencias que controla las características observables y los efectos fijos no observados. A continuación, aplicamos este esquema para medir las repercusiones de los préstamos industriales del Fondo Regional Noreste (FNE) en el crecimiento de empleo y productividad laboral a nivel micro (o de empresa) y el crecimiento del producto interno bruto (PIB) per cápita en niveles macro (municipalidades, micro-regiones y aglomeraciones espaciales) para los periodos 2000–2003 y 2000–2006. Los resultados muestran un efecto positivo y con un significado estadístico de los préstamos industriales del FNE en la creación de empleo a nivel micro pero ningún impacto significativo en el crecimiento del PIB per cápita a nivel macro. Keywords: Impact evaluationRegional development policyFirst-differences methodNortheast regional fund (FNE)Brazil关键词: 影响评估区域发展政策一阶差分法东北部地区基金巴西MOTS CLÉS: Évaluation d'impactPolitique d'aménagement du territoireMéthode de la différence d'ordre 1Fonds régional du Nord-Est (FNE)BrésilSCHLÜSSELWÖRTER: Bewertung von AuswirkungenRegionale EntwicklungspolitikFirst-Difference-MethodeRegionalfonds Nordost (FNE)BrasilienPALABRAS CLAVE: Evaluación de repercusionesPolítica de desarrollo regionalMétodo de primeras diferenciasFondo Regional Noreste (FNE)BrasilJEL classifications: C52R58 Acknowledgements The author thanks Stephen Gibbons, Giordano Mion and Lízia de Figueirêdo, all of whom provided useful comments on earlier drafts and improved the thesis. The author is very grateful to two anonymous referees for constructive and helpful comments; also to the staff of the Institute of Applied Economic Research (Instituto de Pesquisa Econômica Aplicada)/Government of Brazil (IPEA), the Ministry for National Integration (MI, Brazil), and the Bank of the Northeast (BNB), all of which were involved in the regional development funds evaluation project during 2005/2006. This paper benefited from that experience. Previous versions of this paper were presented at the 15th Regional Meeting of the Brazilian Association of Graduate Programs in Economics – ANPEC/BNB – (July 2010) and the 50th European Congress of the Regional Science Association International (August 2010). The views expressed herein are those of the author alone and do not represent the views of the IPEA, the MI or the BNB. The author acknowledges financial support from the IPEA and the Brazilian Federal Agency of Support and Evaluation of Postgraduate Education (CAPES). Notes 1. Evaluation can be defined in several ways: in terms of time (for example, ex ante, mid-term or ex post), levels of complexity (for example, monitoring daily tasks or assessing the impact on the problem), or as an internal or an external evaluation. The definition of Bartik and Bingham (Citation1995) is employed here. It looks at evaluation as a continuum moving from the simplest form of evaluation (monitoring of daily tasks) to the more complex (assessing the impact on the problem). To demonstrate that a programme (or policy) accomplishes its targets, the evaluation must be at the highest level: measuring effectiveness (for instance, it actually does create jobs) or assessing impact (for example, there has been an improvement in the situation). Herein, the terms 'micro-impact evaluation' is used for measuring effectiveness; and 'macro-impact evaluation' is used for assessing the impact on the problem. Also, a cost–benefit analysis needs to be carried out to prove that the programme's benefits outweigh the costs. However, because the data necessary to carry out this cost–benefit analysis are not available, this type of evaluation is left for future research. For further discussion on impact-evaluation strategies, see Khandker et al. (Citation2010). 2. Oliveira and Domingues (Citation2005) employ a municipal dataset to examine whether the Brazilian regional funds (only FNO and FCO are analysed) have a positive impact on regional inequality. The results show that regional growth in Brazil between 1991 and 2000 was not affected by these funds. 3. For details, see Ferreira (2004) and Almeida Junior et al. (Citation2007), who conducted comprehensive studies of the resource allocation each year for these funds. Moreover, it is worth noting that these regional development funds are not the only resources available from a public bank for lagging regions in Brazil. The Brazilian Development Bank (BNDES), a federal public bank established in 1952, also offers loans (interest rates are below market rates but higher than those of the regional funds) to companies of any size and sector in all Brazilian regions. While the focus of the regional funds is the producers in the agricultural sector (60% of total loans), BNDES loans are directed toward large-scale industrial and infrastructure projects (75% of total loans). Moreover, unlike the operator banks of the regional funds that work only in the lagging regions, BNDES addresses the demand for funding in all Brazilian regions and does not have an explicit mandate regarding regional policy. Table A1 in Appendix A compares the loans of regional funds (FNE, FNO and FCO) and BNDES loans by region for the period between 2000 and 2006. 4. For example, see the document on the European Union–Brazil dialogue on regional policy (http://ec.europa.eu/regional_policy/international/pdf/eu_br_regint_en.pdf). 5. For example, see the webpage of the Ministry for National Integration (http://www.integracao.gov.br/fundos/fundos_constitucionais/diretrizes.asp?id=diretrizes) and Banco do Nordeste (2001, 2009). 6. Another proxy for labour productivity at the firm level could be value added per worker; however, this information is not available for the current study. 7. In fact, GDP per worker should be used to indicate average productivity; however, given the lack of accurate data for small municipalities, the total number of workers at the municipal level is not available for most years under analysis. On the other hand, there is a high correlation (the correlation coefficient is +0.997) between total population and total workers at the municipal level in 2000, which is a census year for which reliable information is available. For this reason, the results are not affected by the use of GDP per capita. 8. As suggested by one referee, a computable general equilibrium (CGE) model would have given a wider perspective on the connections between the variables under investigation. The partial equilibrium approach that was applied herein may have limited the current empirical analysis. 9. RAIS data were used under a cooperation agreement between the Labor Ministry and the Instituto de Pesquisa Econômica Aplicada (IPEA). For more details, see Appendix B. 10. Literally, national juridical person registration, as opposed to the CPF number for persons. CNPJ is an identification number for Brazilian companies assigned by the Ministério da Fazenda (Ministry of Revenue). It is comprised of a base of eight digits, a four-digit radical and two check digits, such as 22.222.222/0001-05. 11. Note that information on FNE participation at the firm level is a binary variable given the availability of data. At the macro-level, the available data are a non-binary treatment variable that represents the amount of FNE loans as a ratio of the total GDP at the municipal level. For further discussion on the similarities between binary and non-binary treatments, see Wooldridge (Citation2002, pp. 638–642). 12. At the macro-level, equation (5) can be motivated by the so-called Barro regression. Indeed, many papers that examine the impacts of regional funds on GDP per capita growth – for instance, the studies about European Union Structural Funds, such as Rodríguez-Pose and Fratesi (2004), Dall'Erba and Le Gallo (2008), and Esposti and Bussoletti (Citation2008) – are based on the neoclassical growth model described by Barro and Sala-i-Martin (1991, 1992). Interestingly, Armstrong (Citation2002) discusses some practical steps to reconcile the evaluation evidence on regional policy with the evidence from the growth literature. 13. In other words, Wooldridge (Citation2002) highlights that when D i and (Z 0 i , Z 1 i ) are allowed to be correlated, we need an assumption in order to identify treatment effects. Rosenbaum and Rubin (1983) introduced the following assumption, which they call ignorability of treatment (given observed covariates X i ): Conditional on X i , D i and (Z 0 i , Z 1 i ) are independent. (p. 607) 14. The average annual growth rates are calculated as follows: Z i , g r o w t h = ( ( y i , t / y i , 0 ) ( ˆ 1 / T ) ) − 1 where y i , t and y i ,0, are the final period and the initial period of dependent variable for firm i, respectively; and T is the time period in years. 15. In the micro-analysis, D i ,1999 is a dummy variable that is now zero for all firms in 1999. 16. Angrist and Pischke (Citation2009) argued that since the core assumption underlying causal inference is the same for the two strategies, it's worth asking whether or to what extent matching really differs from regression. Our view is that regression can be motivated as a particular sort of weighted matching estimator, and therefore the differences between regression and matching estimates are unlikely to be of major empirical importance. (p. 69) 17. The robustness check results for the 2000–2006 period show similar qualitative results and, to save space, they are not reported here; they will be provided by the author upon request. 18. To save space, these results are not reported here; they will be provided by the author upon request.