Veja também nosso suplemento literário
A Note on Capital Productivity Using the Brazilian Agribusiness Census
Carlos Feu Alvim*
Claudio David Dimande**
*Chief- Editor of the e&e periodical
**PhD student at PENO/COPPE/UFRJ
For more than 40 years the Brazilian agribusiness census has obtained data on capital stock (K) of the referred institutions. It is probably the only sector for which IBGE has carried out successive evaluations of capital stock. In order to calculate capital productivity (Y/K) it is necessary to know, besides this stock, the aggregated value in these institutions (Y) that are not directly available even though it can be deduced from the published data of these censuses. It is not a simple task to adequately interpret these data and draw conclusions, as we will show.
In order to calculate the aggregated value it is necessary to distinguish the external inputs at the entrepreneurial level or sector, according to the case. Furthermore, sometimes the Brazilian agribusiness census is published with methodological changes that make it difficult to compare results with those of previous years. The individual categories are not always the same, for example, permanent cultures and planted forests (aggregated in 1970 and separated in other years) and machines and instruments and vehicles (aggregated in 2006 and separated in previous years). Other changes refer to the investigated period: in the 1995-1996 Agribusiness Census refers from 08.01.1995 to 07.31.1996 and from 12.31.1995 to 07.31.1996. In this census it is not calculated as well the capital stock and therefore its data are not used in the present note.
The currency updating is another problem as the censuses cover a period when inflation was high and many changes in currency were made. The big difference between the evolution of agriculture prices and the other ones creates problems concerning the correction of products, inputs and agriculture investments values.
For the present study the option was to use currencies of the reference year for capital stock as well as for product and agriculture GDP values. Capital stocks were expressed relative to production values calculated for the institutions in each year. The original values are given in the annex.
The objective of the present note is to make capital productivity approximations using data from the Agribusiness Censuses by substituting the aggregated value for the production value since the aggregated value is not directly available in the agribusiness censuses, but it is in the National Accounting System (agriculture product). It is assumed that the aggregated value obtained from the agribusiness institutions is proportional to those used as proxies.
2. Capital Stock
In Table 1 are indicated the capital stock values expressed in values relative to production values. Values per type of good are shown in an aggregation that permits to evaluate them in all agribusiness censuses considered. Permanent cultures are aggregated to planted forests while machines and equipment to other transport means.
The annual agriculture GDP values are also indicated relative to production value and this value is naturally higher than the aggregated value of the sector (measured by the GDP). However, the comparison of the calculated production value of the institutions with those calculated for the sector is not a direct comparison. Nevertheless, the product values are coherently higher than the GDP of the sector. For 1980 and 1985 the GDP values of the sector relative to production values are much lower than those for the other years. However, it should be remembered that the high inflation in these years may have distorted the values calculation along the year and reduced the profitability in the real economy vis-à-vis financial profits as well.
Properties and Production Values
Figure 1 shows the evolution of capital stock relative to the production value. It is remarkable the large share of the land value in the capital stock that in 2006 represented more than 70% of the total. It should be remembered that, except for careless use such as practices that resulted in soil degradation, it does not undergo depreciation with time, but rather its value is increased when it becomes scarce and since it is used as value reserve its price oscillates with the profitability of other investments and of course of the agribusiness sector itself and this could bias the evaluation.
In Graphic 1 it can be seen that the most important variations of capital stock relative to production value are those due to land value variation. In the same graphic it is used a “transparent” representation of the land value in order to call attention to the relative stability of capital stock (excluding land value) relative to production. In effect, the capital stock value relative to product value has oscillated 10% (standard deviation) around the average value (253/100), as can be seen in Table 1.
Since it is assumed that the production value is a good proxy of the aggregated value, it can be assumed, since the capital stock/aggregated is stable, that the (K/Y) ratio value is stable (reasonably stable along time), and its inverse (Y/K), capital productivity, as well. It should be remembered that the reference is capital stock except land.
Graphic 1: Capital Stock relative to production value
Regarding the “non land” capital stock share, it is noticeable (Graphic 2) the relative stability of the buildings and installations share (about 30%) and that of permanent cultures and planted forests (about 20%). As it should be expected, with the drastic modernization along these 36 years, the share of machines and equipment has grown (doubled) and that of livestock and animal traction has decreased.
Graphic 2: Capital Stock excluding the land factor.
3- Capital Productivity (Proxy Values)
As an approximation to the behavior of values aggregated by the enterprises, this notes uses: a) the production values of the enterprises included in the Census and b) the Agribusiness GDP.
The capital stock is considered with and without the land value. The values shown in Table 2 are obtained by dividing the data of the two last lines (Y proxies) of Table 1 by the value of the first line (K). In the following lines are indicated the values resulting from the division of capital stock, excluding the land value (line 6 of Table 1), by the values of the last two lines of Table 1). It is also indicated the average values for the considered censuses.
Table 2: Production Value or GDP per capital stock
(with and without land)
In order to estimate the capital productivity value it is necessary to use a measure of the product that expresses the aggregated value such as the agribusiness GDP. It should be noticed that the sum of the aggregated values of all institution is different from the aggregated value of the sector since the limits of the sectorial calculation and those of the institutions are different. Regarding the capital stock sum of all institutions of the country, it coincides with the national capital stock. The agribusiness GDP value (of the national accounts) / Census capital stock ratio can be considered as a good estimate of the capital productivity value.
The capital productivity value, including the land value, is low (0.08) relative to the Brazilian average capital productivity that is around 0.40 (see the article “Capital Productivity Indexes in the Brazilian Agribusiness” in this issue). It should be noticed that the land value is not directly included in the Gross Formation of Fixed Capital that is the basis for determining the stock using data from the National Accounts. For comparison purposes with other sectors, it is better to consider the capital stock without land whose average value is 0.26. This value is lower than the average of the economy as indicated above. Since we are considering a primary production sector that starts an aggregation chain with steps of higher productivity, the capital productivity value in the agriculture sector can be considered as a primary activity.
Concerning the behavior along time, the product value has the advantage of originating in the same base and expressed in the same currency. Graphic 3 shows the evolution of different factors relative to the 1970 value and the different ratios indicated for capital productivity evaluation (from Table 2). As already mentioned, it is expected that the behavior of the production value / capital stock ratio is a good approximation for the capital productivity evolution along the years.
Graphic 3 – Capital Productivity Indexes
The most significant index of capital productivity seems to be the agribusiness GDP / Capital Stock (without land) value. This index shows an agribusiness capital productivity that is relatively stable (upper green and continuous line in Graphic 3). This behavior is similar to that shown in “Capital Productivity Indexes in the Brazilian Agribusiness” in this issue where it was concluded that the aggregated value/hectare ratio has remained practically constant for various decade while, at constant prices, there was a high growth of the agribusiness production per ha and of capital productivity.
The published censuses present important information, namely capital stock subdivided in different items. The most variable component of this stock is the value assigned to land. When the other components are considered, there is a noticeable constancy in capital stock relative to production value. The share of the different components in the stock is relatively stable and it is only noted a strong increase of machines and vehicles share in detriment to livestock and traction animals.
The capital productivity behavior along time does not show a growth or decrease trend at current values. As it is known that there is an accumulated delay that caused a 60% drop in the agribusiness prices relative to the other ones, it can be concluded that there was a strong increase of agriculture productivity at constant prices that is not perceived at current prices.
5- Bibliographic References.
INSTITUTO BRASILEIRO DE GEOGRAFIA E ESTATÍSTICA. 2006. Censo Agropecuário – Brasil. Rio de Janeiro: IBGE.
INSTITUTO BRASILEIRO DE GEOGRAFIA E ESTATÍSTICA. 1995/1996. Censo Agropecuário – Brasil. Rio de Janeiro: IBGE.
INSTITUTO BRASILEIRO DE GEOGRAFIA E ESTATÍSTICA. 1985. Censo Agropecuário – Brasil. Rio de Janeiro: IBGE.
INSTITUTO BRASILEIRO DE GEOGRAFIA E ESTATÍSTICA. 1980. Censo Agropecuário – Brasil. Rio de Janeiro: IBGE.
INSTITUTO BRASILEIRO DE GEOGRAFIA E ESTATÍSTICA. 1975. Censo Agropecuário – Brasil. Rio de Janeiro: IBGE.
INSTITUTO BRASILEIRO DE GEOGRAFIA E ESTATÍSTICA. 1970. Censo Agropecuário – Brasil. Rio de Janeiro: IBGE.
Annex 1: Capital stock values.
Annex 2 – Production and GDP value in the currency of the time
Graphic Edition/Edição Gráfica:
Friday, 06 January 2012.