The Quality of Investments

In January, President Lula announced a Growth Acceleration Plan (PAC, for its acronym in Portuguese) for the four year period 2007-2010. The PAC is a group of measures intended to increase infrastructure investments, and through these investments, foster growth. Of the roughly BRL 500 billion announced, more than half are in energy (thermo and hydro electric generation plants, transmission lines and gas pipelines), another third are in social infrastructure (water supply and sanitation, housing) and just over 10% in logistics (mostly roads, railways and ports). Some BRL 70 billion of the BRL 500 billion are federal public investments, another BRL 235 billion are state company investments and BNDES (Brazil’s development bank) is expected to have a heavy role in financing a good chunk of the remaining investments expected from the private sector. If the quality of these investments – the extent to which they actually impact growth and productivity – depend on the existing incentives behind the good use of public resources and accountability for results, there may be reason for concern.  

First, consider public and state enterprise investments. One paper done last year by IPEA (Brazil’s governmental economic studies institute) compared public investments in Brazil, Argentina and Chile during the period 1970-2000, and concluded that in all three countries public investments were correlated with an increased product. However, only in Chile were public investments correlated with increased productivity. In addition, in Brazil there was some indication of a crowding out effect of private investments. Vito Tanzi has an interesting (even if a little old) 1997 IMF Working Paper (with Hamid Davoodi) where some evidence is provided for a correlation between greater public investment, greater corruption and lower quality of public infrastructure. The model describes corruption as inducing larger investment budgets with reduced productivity of the investment projects and lower budget resources for operations and maintenance. The same results need not derive from bribes but would follow by politicians leading public investments to regions of their constituency or any other situation where the productivity of public investments is diminished by lack of adequate controls.  

Accountability for bad public investments is pretty much non-existent in Brazil. In addition to the impunity generated by the poorly functioning justice system and corruption, there seems to be little knowledge of the actual impact of public sector projects. In 2005, Brazil created an inter-ministerial Monitoring and Evaluation Committee to oversee monitoring and evaluation of all projects included in Brazil’s four year plans. This committee must approve these projects before sending them to Congress. For any large projects proposed by government ministries (over BRL 10 million) as well as state enterprises (over BRL 67.5 million) the committee requires feasibility and socioeconomic studies of the proposed projects. The PAC 2008-2010 investments of the federal government and state enterprises depend on the approval for the 2008-2011 four year plan that goes to Congress this year by August 31. This seems to me an important step towards greater accountability. However, at present, the monitoring of the projects included in the four year plans is limited to information inputted by the ministries and state enterprises themselves in an automated system with no external verification and very few (if any) impact evaluations.  

Second, consider Brazil’s BNDES. I have recently learned that information on BNDES operations is surprisingly not easy to get a hold of. Since 2005, a group of NGOs has been requesting greater access to information on BNDES operations apparently interested in the social and environmental impacts of its private sector loans. A relatively simple request for a list of the 10 largest loans in each sector where the bank has been active has been denied and is apparently non-public information. At some point, in response to this pressure, BNDES then president (and now Minister of Finance) Guido Mantega set instructions for the establishment of a clear information policy by the bank, noting that there were also demands for greater transparency from other state institutions such as the Congress and Public Ministry. The information policy seems to never have been developed.  

The importance of accountability in the use of BNDES resources is enhanced by the volume and nature of its operations. During Lula’s first mandate, BNDES annual disbursements rose from the roughly BRL 30 billion they had been in 2003 to about BRL 50 billion in 2006 and this year they are expected to reach some BRL 60 billion, doubling disbursements from the pre-Lula era. A couple of weeks ago, the recently appointed president of the BNDES announced a new “industrial policy” in which the bank will identify specific groups in specific sectors that it plans to “strengthen” so they can compete internationally (sectors cited were mining, metallurgy and orange juice); it will assist multinational groups in Brazil to modernize their plants to compete with other countries for production of new products being launched by those multinational companies (e.g. in the automobile industry); and where it will assist some labor intensive industries to move to greater value added products in sector such as footwear and apparel. This exercise in “picking winners” generally leads to rents and investments of low productivity and even a recent defense of such efforts (Rodrik and Hausmann) recognizes the importance of a proper institutional setting (open and transparent) for any chances of success.  

Brazil’s effort to increase public and private investments are very welcome, particularly since gross fixed capital formation seems to have been stuck at around 16-17% of GDP for the last decade and so far, this year, under expectation of greater tax revenues, the federal government seems to be increasing its consumption even more than its investment. However, with the possible exception of the ongoing discussions on environmental licensing requirements for infrastructure, I have seen very little attention, if any, from the private or public sectors – or from the media – on making sure these investments are made in the best of public interest.