Economy & Energy
Year VIII -No 45:
August-September 2004  
ISSN 1518-2932

seta.gif (5908 bytes)No 45 Em Português






other e&e issues

e&e No 45

Main Page

The technological prospective prediction with a simple mathematical modeling

More Work and Jobs with the same Capital or How to Increase the Capital Productivity

No 45 in Adobe pdf for download

e&e links



Text for Discussion:


Carlos Feu Alvim


1 – Introduction

Brazil presents low capital productivity relative to its development level, as we have shown in previous articles.

Efforts aiming at increasing productivity in Brazil have almost always been focused on the abundant input (labor) instead of on the scarce one (capital). It is not rare as well to relate productivity to physical production and not to aggregated value. Now if we take the example of an industry (capital intensive) such as the steel industry, it makes no sense at all to evaluate the productivity in tons of steel per laborer if it is not taken into account the commercial value of the product and the efficiency in using the capital.

A development policy for Brazil should emphasize the productivity of existing capital goods and those to be installed. In our evaluation, this could significantly contribute to overcome the limitations to growth that Brazil has been facing in the last 25 years. In fact, each percent point gained in the global capital productivity means – for the same capital stock – the same percent growth in the GDP.

Therefore the challenge is to generate more product and more jobs with the same capital. Offering recommendations for such a policy –like those adopted in other countries such as England, Australia, New Zealand and Germany – is the objective of the present paper.

In the present article we will: (1) recall the macroeconomic diagnosis already made, (2) comment on the sectorial (microeconomic) diagnosis of the Brazilian industries carried out by the McKinsey Consulting company that has already performed other comparative studies at the international level, (3) identify the needs regarding understanding the subject and (4) suggest some macro- and microeconomic measures aiming at increasing the capital productivity.


2 - Recalling the Macro Diagnosis

From 1999 on Brazil needs twice the capital that was necessary in 1970 to generate one unit of product and almost three times that of 1950. This larger capital quantity for generating one product unit, together with investment decrease, has been halting our development. A developing country where capital is the scarce factor should have a larger capital productivity than that of the developed countries. This productivity would gradually decrease along its normal growth process.

However when this productivity swiftly approaches the level of the developed countries without the corresponding income of those countries (as is the case of Brazil) we are facing a strangulation point in the normal growth process of the country.

Concern regarding the capital productivity decrease can be identified in Brazilian studies where the capital/product ratio (K/Y) appears as an end or a means for determining growth scenarios. Besides our previously mentioned papers, we could mention papers published by the Instituto Nacional de Pesquisa Aplicada – IPEA (National Institute of Applied Research), such as Morandi, Zygielszyper, Reis (2000), Bacha and Bonelli (2001) and by the Banco Nacional de Desenvolvimento – BNDES (National Development Bank), Carvalho (1996). We could also mention papers that examined the problem in Latin America, Hofman(1992) and Hofman (1997).

In Aumara Feu’s thesis (2003) about capital productivity we can distinguish three factors that, from the macroeconomic point of view, have contributed to reducing capital productivity:


a) Content effect – increase of capital stock by product unit in the sectors;

b) Structure effect- reallocation of production to sectors that are more capital intensive;

c)    Price effect– price increase of capital goods relative to the price of other products.


For the purpose of the present paper it is important to distinguish the contribution of each effect. It is also interesting to consider our productivity at international prices in order to evaluate our competition relative to other countries. In Figure 1 it is shown the evolution of the capital/product ratio (a) estimated in current values (products and investments corrected by the GDP deflator), (b) at constant prices and (3) at international prices.


Comparing the evolution in constant values and in current values one can evaluate the price effect. The capital productivity between 1970 and 2000 has dropped 455 points in current values. Out of this decrease, ten points can be ascribed to the price factor. One can see in Figure 1 that price had a preponderant role in the second half of the eighties.


A distinction between the content and structure effects was established in the mentioned thesis. For the developed (OECD) countries it was possible to evaluate the two effects along two and a half decades (1970 –1994) since sectorial investment data was available. For Brazil it was only possible to analyze (in the mentioned thesis) the 1985-1994 period using IBGE’s sectorial structure and the sectorial capital product ratios calculated for OECD countries. From this evaluation it resulted that ¾ of the productivity decrease in that period (at constant prices) can be ascribed to the structure effect[1]. That is, Brazil has changed its product distribution among the several sectors by privileging the capital intensive ones.


Figure 1: K/Y ratios in Brazil, with current price and constant price investment in national and international currency. Source: Aumara Feu’s   thesis (2002)

It should be recalled in the present recapitulation the results obtained for the capital/product ratio in what concerns capital stock of the following items: machines and equipment, others, residential civil construction and non-residential civil construction.

Contrary to what a hasty analysis could conclude, the ratio between the stock of machines and equipment and the total product has not changed along time. It reached a maximum in 1980 and has been decreasing since then. Incidentally, in absolute value, the stock of machines and equipment at the beginning of the eighties has decreased in the middle of the nineties and was resumed at the beginning of 2000. That is, the real value of our stock of machines and equipment is now almost equal to that of 1980.



Figure 2: Contribution to the Capital / Product ratio from each type of good. Source: e&e No 44.

Therefore, we can infer from Figure 2 that the non-residential constructions segment[2] has the largest responsibility regarding capital productivity decrease (increase of the capital/product ratio).

We can conclude in this recapitulation that:


· Capital productivity in Brazil is low, taking into account its development level;

· The price of capital goods is responsible for about 20% of capital productivity decrease (10% in 45%);

· There are indications that part of the productivity decrease was due to the fact that productivity was shifted to capital intensive factors;

· The main responsibility regarding capital productivity decrease falls on the increase of non residential construction goods relative to the GDP.


3 – Micro Diagnosis

There are studies comparing the same productive sectors at the microeconomic level in different countries in order to make a diagnosis of the real situation of the factor’s productivity in production units.

These studies point out management techniques so that the resources would be allocated in a productive and efficient form. The consulting company McKinsey has carried out one of its comparative studies for Brazil. The results for some sectors are shown in Table 1.

Total Productivity in the Studied Industries (USA=100)








Capital share(2)
















Car Industries (1)





Source: MCKinsey (1) only passenger cars, (2) estimated using Cobb-Douglas function with capital share estimated for a set of studied countries. [3]

The results of the studies show a capital productivity 20% lower than that of the USA for three of the four industries presented and higher than that of that country in the car industry.

In order to compare the industrial level diagnosis with the macroeconomic one it is necessary a compatibility of parameters between both type of studies which naturally does not exist[4]. However, as the sectorial analysis uses homogeneous criteria for the different analyzed countries, it is possible to use the results of the sectorial studies in order to favor a macro objective that would consist in increasing the capital productivity of the country.

The McKinsey study – as well as others that might exist in this area – could and should be used to reach the global objective. From the preliminary analysis of its executive summary (the sectorial studies are also available at the Internet) one could arrive at the (hasty) conclusion that there is not much to be done in the capital productivity area since the difference between our productivity and that of the United States is not higher than 25% in none of the sectors[5]. There are much higher gains to be obtained in the labor productivity. However, we would be forgetting that at the same development level in which we are now, the capital productive of the countries that now belong to the first world was much higher than it is now.

Let’s take, for example, the automotive sector where the capital productivity was – as was expected in a developing country – 70% higher than the American one (in 1994). Probably, not by chance, the labor productivity in this sector (pointed out in the study) was only 31% of the American corresponding value.

A more rational criterion to increase the productivity would be to increase the total productivity (relative to that of the USA). However, when this is done, one should pay attention to the remuneration division between capital and labor[6]. In the McKinsey study, due to the comparative objective, the remuneration division is based on the average value of the countries under study, namely 30% for the capital and 70% for the labor in the case of the car industry. But this remuneration division is valid for the analyzed countries while it is certainly not valid for Brazil, where the labor remuneration is generally lower. That is, even when it is intended to increase the total productivity relative to that of the USA, one may not be taking into account that in Brazil the capital weight is greater than in the USA. Perhaps it would be more interesting to maintain the existing comparative advantage instead of increasing the labor productivity in detriment of that of the capital.

It should be emphasized that the McKinsey study does not limit itself to suggest measures for one or other specific sector but it outlines a general framework for the different sectors. The study tries to quantify the job losses and the investment that would be necessary for the surplus labor to be absorbed by the resulting growth and, besides, to increase the formal jobs. According to the study, with  5% growth per year there would not be a decrease in informal occupation due to the increase of the projected labor productivity in Brazil. It would be necessary to increase the present investment rate from 19% to 26% of the GDP and grow 8.5% per year during ten years in order to decrease its informal jobs from 50% to 40%. Furthermore, the country would have to increase its exports in about 12% annually in order to face the necessary imports to reformulate and modernize its productive park.

That is, the total productivity increase aiming at reaching the labor productivity of the developed countries makes macroeconomic sense only in an accelerated economic growth framework. Without that growth, unemployment and informality would be aggravated.

4 – It is Necessary to Deepen the Diagnosis

Initially it would be necessary to deepen the diagnosis concerning the causes of inefficiency in the use of capital goods stock. A wrong diagnosis could lead to measures that would aggravate the problem instead of alleviating it. For example, Brazil has presented in the last years a low investment rate (gross formation of fixed capital). The credit offer, with subsidized interests, or even the Government’s guarantee, through administered prices, regarding investments remuneration may aggravate the problem of low capital productivity[7].

As we have pointed out, the capital productivity decrease (at constant prices) may be caused both by directing production to techniques that are more capital intensive and by reallocating capital to sectors that are more intensive in using this factor. In the real economy, where relative prices vary, capital goods prices is the other important factor.

Therefore, knowing the behavior of the variables that affect the capital productivity at the aggregated and non-aggregated levels might indicate measures at the macroeconomic and sectorial level to improve the performance of the country regarding these items.

Encouragement should be provided to studies concerning private and government planning aiming at orienting investments to sectors or techniques that are less intensive in using that factor and generating more significant growth scenario for the future.

The preliminary items of the policy to be adopted in order to overcome the present situation would be removing the obstacle in the productive system, specially the process factors, and concern regarding the quality of investments and not only its quantity.

Measures for implementing the capital productivity

Increasing the capital productivity of a continental country like Brazil is not an elementary task,. mainly when one does not have a definitive diagnosis. In what follows, we list some measures, both global and local, that could, in a preliminarily analysis, contribute to increase the capital productivity:

1.    Larger use of the productive park: an important measure in the reduction of the capital/productivity ratio is to make possible the more intensive use of the existing production goods, amplifying the working days and hours. This will have as consequence an increase in work demand and job offer using the same productive park. To materialize this objective would require, for example, measures that would favor, without any loss to the laborer, the definitive or temporary elimination of labor limitations to the larger use of the productive park.

2.     Give priority to productive sectors that are less capital intensive: an analysis of the capital productivity could be a tool for granting favored credits[8].  Evidently, incentives to investment should not result in stimulating inefficiency regarding capital use and consequently capital/product ratio decrease.

3.    Motivate capital productivity sector-wise: one sector can improve its productivity by better using the production means. Like it is made in quality programs, measures in the management area could increase, at the same time, both capital and labor productivities.

4.     Increase in a qualitative way the aggregated value in each sector: since the capital productivity corresponds to the product/capital ratio, it can be increased by reducing the denominator or increasing the numerator. The numerator could be increased if the country would try to participate in more advanced stages of production and not in the intermediate products (predominantly, our exports line). Other forms for valorizing our product, such as improving its quality, its sophistication level and incorporating new technologies should also be pursued. One should not forget the commercial aspects such as establishing trade marks that constitute a significant part of the aggregated value. [9]

5.    Coherent taxation aiming at increasing capital productivity: in a phase where it is necessary to increase the production capacity, it is preferable to tax consuming goods instead of production ones. Investment stimulation must be selective to prevent suppression of jobs, which cannot be economically justified without subside to capital.

6.    Promote development in medium and small cities instead of large urban areas: in the analysis it was concluded that the large capital/productivity ratio (decrease of the capital productivity) is connected with non-residential constructions that characterize the large human concentrations. Encouraging urban dilution means to reduce the need to invest in urban structure by generated product.

7.    Reduction of costs associated with the Government: It is necessary to reduce the investment costs associated with the bureaucratic, regulatory and administrative processes. Reduce the duration of works that, due to bad planning or multiplicity, permits the capital to remain idle for several years, increasing the capital/product ratio and charging investment costs. Mainly in the public sector, reduce costs through the rationalization and moralization of the public bidding process.


We started with the verification that the capital productivity decrease is one of the reasons of the stagnation of the growth per capita in Brazil. In the present article, the idea is to draw up a schedule using a diagnosis about the capital productivity decrease in Brazil and indicate possible actions aiming at increasing the capital productivity and consequently increase growth in the country.

Identifying the measures is neither exhausting nor definitive. It would not be surprising that a deeper analysis would indicate that some of them are inefficient or even disadvantageous.

If the subject is relevant, and we believe so, it should be considered and analyzed by society, including class, governmental and legal organizations. It would be desirable that this process would not retard actions whenever there is a reasonable consensus about the subject.

It should be recalled that some actions are already under way and in some of them the governmental contribution may be just not to upset it. In the previous issue of our periodical we have mentioned some promising facts such as night flights with promotional prices that favor the larger use of the capital invested in airplanes and airports and the fact that the Ford company uses its production capacity during 24 hours of the day in some of its units. Recently the subway company in Rio de Janeiro has offered cheaper tickets for early passengers. Shopping centers have also realized that in order to compensate investments it was necessary to be open during more hours. On the other hand, even though for reasons different from those pointed out, it is already observed a trend regarding the transfer of producing centers to smaller urban centers.

In other points it seems that we are going in the direction opposite to the desired one: particularly the taxing policy and even (in some points) the labor policy, namely favoring capital in detriment of labor. Points of concerns are the additional costs imposed on the productive investments due to legislation, regulation and bureaucratic excesses. The lack of planning, the economic uncertainties and the excessive interest rates also continue to delay and make investments more costly, reducing the capital productivity


BACHA, Edmar L. e Regis Bonnelli, ”Crescimento e Produtividade no Brasil: o que nos diz o Registro de Longo Prazo.” Rio de Janeiro: Seminários da Diretoria de Estudos Macroeconômicos do IPEA 52, 2001.

CARVALHO, José Carlos, “Estimativas do Produto Potencial, Relação Capital/Produto e Depreciação do Estoque de Capital.” Rio de Janeiro: Textos para Discussão, Área de Planejamento do Departamento Econômico do BNDES 44, 1996.

FEU ALVIM, Carlos. “Capital Productivity: A further Limitation for the Brazilian Growth” Economy & Energy, 44 (June-July), 2004.

FEU, Aumara. Produtividade do Capital no Brasil de 1950 a 2002. 151 f. Tese (Doutorado em Economia) - Universidade de Brasília, Brasília, 2003.

______. “ Evaluation of Capital Productivity in Brazil in the XX Century.” Economy and Energy. 43 (May-June). 2004

HOFMAN, André. Capital accumulation in Latin America: a six country comparison for 1950-1989. Review of Income and Wealth, v. 38, n. 4, p. 365-401, Dez. 1992.

______. The economic development of Latin America in the twentieth century. Northampton, MA: Edward Elgar, 2000.

MORANDI, Lucilene; ZYGIELSZYPER, Nora; REIS, Eustáquio. Tendências da Relação Capital/Produto na Economia Brasileira. Rio de Janeiro: IPEA, oct. 2000. (Boletim Conjuntural do IPEA n. 51)


[1] It was carried out an analysis of the intensity effect evolution in two sectors in Brazil, for which investment data was available. The conclusion was that the intensity was underestimated, at least for these sectors.

[2] One of the factors that has contributed was the fast urbanization process in Brazil and the growth of large agglomerations. Between 1940 and 2000, the urbanization rate in Brazil grew from 31% to 81%. The urban infrastructure can be in part responsible for the capital productivity reduction.

[3] The total productivity here is the geometrically weighted average of productivity (capital and labor) relative to those of the USA. It is not the so called “total productivity of factors” commonly associated with technology.

[4] The study points out an economical productivity as a whole that is 60% higher than that of the USA, what contradicts the conclusion of our paper and those of other authors who studied the subject. Adopting depreciation curves similar to those used for accounting purposes (shorter amortization period) would lead to such results.

[5] In telecommunication, where it is 75% relative to that of the USA.

[6] The capital remuneration in developing countries like Brazil is expected to be higher than that of central countries. This remuneration would imply a higher weight for the capital productivity when the total productivity is calculated.

[7] The fact that elections will soon be held (the present month is July 2004) has prevented (or postponed) increasing contributions of enterprises to Social Security. Government authorities have promised that there would be a compensation by tax reduction in other areas. According to what was later revealed, these measures were intended to reduce investment costs (capital) that would compensate a cost imposed on labor.

[8]  It should be noted that capital productivity evaluation is not intuitive. For example, public service is capital intensive since large investments in civil works per public servant are necessary and the servant’s salary is the element used for calculating the contribution of public service to the GDP. The same is true in what concerns tourism, that requires large investment per product.

[9] In spite of the fact that the footwear industry in Brazil is competitive, it has no important trade marks. Brazilian products are frequently sold abroad using foreign trade marks that hides their origin.




Graphic Edition/Edição Gráfica:
Editoração Eletrônic

Tuesday, 11 November 2008

Contador de visitas