A longstanding question in economics is why some countries are much wealthier than others. Recent studies show that one part of the answer has much to do with differences in productivity levels among countries. Since the seminal work of Solow (1), an important concern in economic literature has been how to measure productivity. If productivity has a strong influence on growth and welfare in the long run, having a good productivity measure is a crucial element.  

Several studies on development revealed that the gap between Latin America and the developed world has increased. According to Maddison (2) the income per capita in Latin America was, on average, 25% of the income per capita in United States (U.S.) between 1950 and 1980, and it decreased to 20% between 1980 and 2000. Following this, Cavalcanti, Pessoa and Veloso (3), found that Latin America’s Total Factor Productivity (TFP), relative to the U.S., decreased over time. Indeed, Latin America’s TFP was around 90% from 1960 to 1975, but then fell to 62% in 2000. They came to these results by performing a modified growth accounting exercise which includes physical and human capital data. 

Another study, performed by De Gregorio (4), arrives at similar conclusions. He found that, in 2000, the purchasing poverty parity corrected GDP per capita in Latin America has been only 21% of the U.S. GDP. He suggests that the main element behind this issue is the difference in productivity among Latin America and U.S. He came to that conclusion by taking into account that in 2000 the capital/output ratio in Latin America was 27% less than that of the U.S., the accessibility to qualified human capital was 42% (5) less, but the TFP was 57% lower. Certainly, there are more differences in the performance of the production factors than in the factors availability. Then, it seems that policies aimed to promote a convergence process in terms of per capita income have to focus mainly in productivity.  

If productivity is so important, then it is necessary to find the best method to compute it. It is widely accepted that the best way to compute productivity is at the micro level, specifically, at the firm level. It can be done by calculating the output/employment and output/capital ratios. But, for these calculations, the best scenario is one where output and inputs are available in physical terms (i.e. quantities produced, man-hours, machine-hours, etc). However, this scenario greatly depends on availability of information collected in enterprise surveys and usually these datasets do not contain this kind of information. 

To deal with this problem economists have used aggregate variables for the computation of these ratios and in particular for the computation of TFP. Growth accounting exercises became popular in the macro literature, and TFP measures proliferate in particular for developed countries. But these measures do not only face the aggregation problem of variables, but also a problem that is typical in macroeconomics: TFP is computed as a residual, and then, nothing guarantees that this residual is reflecting purely productivity.
 
Nowadays, the literature on productivity has turned to evaluating TFP using firm-level data. In other words, it has gone from macro to micro again. Nevertheless, the data problem is still present and in the absence of physical information on capital and labor, economists have to use nominal variables, i.e. value-added, book value of capital remunerations, among others. The problem with these revenue based measures of TFP is that they could imply serious misspecifications problems, as TFP summarizes not only pure productivity, but other factors like change in prices, economic cycles, market power, public policies, and many others that may cause large distortions on the results. 

Nevertheless, new measures of TFP have been developed since the Restuccia and Rogerson (6) and Hsieh and Klenow (7) papers, as TFP is a combination of revenue productivity and a physical productivity. This new literature stresses the fact that productivity is low due to misallocation of resources. Economics is the science which studies human behaviour as a relationship between ends and scarce means which have alternative uses. Scarcity means that available resources are insufficient to satisfy all wants and needs. The subject thus defined involves the study of choices in the allocation of resources as they are affected by incentives, barriers and availability. 

Economics started as a science where the main subject was the allocation of resources. In the beginning this allocation was conceived at a micro level. The rise of Keynesian economics created a place for macroeconomics as the other main branch of economic theory along with microeconomic theory (8). The fact that economists are studying again productivity in a micro context reveals that in the twenty-first century economists are trying to clarify the connection between the principles of macroeconomics and the more familiar principles of microeconomic theory.