This Selected Issues paper and Statistical Appendix analyzes economic developments in Botswana during the 1990s. The paper analyzes the growth process during 1982/83–1996/97 by assessing the contribution of capital, labor, and technological progress, both at the macroeconomic and sectoral levels. The paper examines the diversification initiatives undertaken by Botswana and the extent to which diversification and employment creation have been achieved. It provides the background to the unemployment problem, and summarizes Botswana’s policy initiatives to diversify and create sustainable employment.

Abstract

This Selected Issues paper and Statistical Appendix analyzes economic developments in Botswana during the 1990s. The paper analyzes the growth process during 1982/83–1996/97 by assessing the contribution of capital, labor, and technological progress, both at the macroeconomic and sectoral levels. The paper examines the diversification initiatives undertaken by Botswana and the extent to which diversification and employment creation have been achieved. It provides the background to the unemployment problem, and summarizes Botswana’s policy initiatives to diversify and create sustainable employment.

II. Sources of Economic Growth29

A. Introduction

41. Botswana experienced impressive growth over the past two decades, averaging 11 percent during 1982/83-1989/99 (July-June), before moderating to 5 percent during 1990/91-1996/97 with the slowing of the expansion of the mining sector. A key challenge in the period ahead is to maintain this strong economic performance by achieving growth rates consistently higher than the rate of population growth, thereby raising per capita income and reducing poverty.30 To better understand how this objective might be achieved, this paper analyzes the growth process during the 1982/83-1996/97 period by assessing the contribution of capital, labor, and technological progress, both at the macroeconomic and sectoral levels.

42. An important aspect in examining Botswana’s recent growth performance is to determine whether the growth process has been intensive or extensive, where intensive growth denotes technology/efficiency-driven growth and extensive growth is achieved by employing more factor inputs. This distinction is important because, with intensive growth, a higher growth rate can be sustained with the same factor inputs. The remainder of this paper is organized as follows. Section B discusses the growth accounting framework, Section C analyzes the sources of aggregate and sectoral growth, and Section D provides a cross-country comparison of total factor productivity (TFP) and marginal productivity of capital. Section E analyzes the effects of improving the quality of factors of production on TFP, while Section F presents some preliminary conclusions.

B. Growth Accounting Framework

43. The growth accounting framework is based on a Cobb-Douglas production function, defined as:

Yt=AtKtαLt1α,(1)

where Y is output, A is a technological parameter, K is the amount of capital used, L is the amount of labor used, and a is the factor share parameter, with a value between 0 and 1.

From equation (1), it can be derived how output increases over time:

ΔYY=ΔAA+αΔKK+(1α)ΔLL(2)

Equation (2) states that the growth rate of output is composed of the growth rate of technological progress in total factor productivity (TFP), the growth rate of capital stock, and the growth rate of labor supply, where the capital and labor factors are weighted by their shares in the production process. The growth equation also leads to an interpretation that explains long-run growth. Given that effective labor supply is bounded, and higher capital growth than labor growth would lead to diminishing returns to capital, higher growth can only be achieved by increasing TFP. The estimate of TFP is derived from equation (2) as:

ΔAA=ΔYYαΔKK(1α)ΔLL(3)

44. The key variables used to estimate TFP are therefore capital stock, labor supply, output, and factor shares. There are two main methods used to estimate factor shares. The first method relies on the national accounts statistics, where factor shares are estimated by measuring the share of income that is distributed to each factor of production. This methodology, however, suffers from a number of weaknesses, including the lack of data on compensation to workers (other than formal sector employees) and its implicit assumption of perfectly competitive capital and labor markets. The second approach relies on estimates of factor shares calculated by regressing the growth rate of output on the growth rate of each input and a constant, where the estimate of each factor share is the estimated coefficient of the relevant input and the constant in the regression can be viewed as an estimate of the growth rate of TFP, This methodology, unfortunately, also suffers from a number of weaknesses, including the assumption that factor shares remain constant over time or across countries.31 Estimates of TFP, however, are quite sensitive to changes in the value of capital shares (α).

45. To avoid these problems, Sarel (1997) estimates capital shares in five Association of South East Asian Nations (ASEAN) countries, using internationally comparable data from the Summers-Heston database to estimate capital and labor factor shares and assuming that technological factor shares are determined by the industrial structure of the economy and, possibly, its level of development. A comparison of Sarel’s sectoral shares and those derived from the 1992/93 Social Accounting Matrix (SAM) of Botswana shows some differences (Table II.1). This study relies on shares derived from the 1992/93 SAM.

Table II.1.

Botswana: Capital Shares in Botswana and ASEAN Countries

article image
Sources: Sarel (1997); and Fund staff estimates.

Derived from the 1992/93 Social Accounting Matrix.

C. Sources of Aggregate and Sectoral Growth

46. This section summarizes the results of applying the growth accounting framework discussed above to Botswana, Growth is attributed to changes in factors of production (capital and labor) and TFP for four distinct periods during 1982/83-1996/97, both at sectoral (agriculture, mining, and manufacturing) and aggregate levels.

Overall economy

47. Botswana sustained high economic growth during the 1980s, averaging 11 percent per annum (Table II.2). During the first half of the decade, growth was mainly driven by capital accumulation, which averaged 6.7 percent a year, and a strong growth in employment, which averaged 3.2 percent a year. Technological progress during this period, however, was slow, with TFP growing by only 1 percent on average per year. The subsequent period, 1986/87-1989/90, witnessed a strong improvement in TFP, which grew at an annual average rate of 5.6 percent. This improvement, combined with strong growth of capital (averaging 13.3 percent a year), led to significantly higher average output growth of 12.8 percent.

Table II.2.

Botswana: Aggregate Growth of Capital, Labor, and TFP, 1982/83-1996/97 1/

(Average annual percentage change)

article image
Source: Fund staff estimates.

National accounts year beginning April 1.

48. Output growth moderated during 1990/91-1993/94 to 4.8 percent when TFP growth stagnated, and the growth of capital, although still high, slowed to 9.5 percent. During the most recent period (1994/95-1996/97), overall growth picked up to 5.5 percent a year, largely on account of sustained growth in factor inputs. During 1990/91-1996/97, the growth of labor inputs at 4 percent a year helped to moderate the slowdown in output growth.

Mining

49. At independence in 1966, the mineral industry’s contribution to GDP was negligible. Since the discovery of diamonds in the late 1960s, mining operations have continued to expand. The mining sector now contributes more than one-third of real GDP, with diamond mining accounting for more than 90 percent of all the mining activities.32 In value terms, Botswana is the largest producer of diamonds in the world. The rapid growth of diamond mining is the primary factor behind Botswana’s impressive growth in real per capita output, which increased from US$572 in 1971 to US$3,278 in 1996. Diamonds also contribute about 75 percent of Botswana’s export earnings and more than 45 percent of central government revenue. However, given the highly capital-intensive production process, the mining sector provides less than 4 percent of formal sector employment, and annual labor growth in the sector averaged only 1-2 percent during the 1982-97 period, despite strong output growth over the period (Table II.3).

Table II.3.

Botswana: Growth of Capital, Labor, and TFP in Mining Sector, 1982/83-1996/97 1/

(Average annual percentage change)

article image
Source: Fund staff estimates.

National accounts year beginning April 1.

50. During 1982/83-1985/86, the mining sector grew by an impressive 17 percent per year, mainly driven by the coming onstream of new capacity, which was reflected in sizable TFP growth, averaging 19 percent per year. Meanwhile, employment growth was weak, and the capital stock fell by 3 percent annually (Figure II.1). During the 1986/87-1989/90 period, output growth in the mining sector slowed to more sustainable levels (6½ percent per year), notwithstanding moderate growth in capital and a modest pickup in labor growth. The sharp decline in TFP growth during 1986/87-1989/90 was the main explanation for the deceleration of output growth.

Figure II.1.
Figure II.1.

Botswana: Output and Productivity, 1982/83-1996/97 1/

(Annual percentage change)

Citation: IMF Staff Country Reports 1999, 132; 10.5089/9781451806342.002.A002

Sources: Botswana authorities; and Fund staff estimates.1/ National accounts year beginning July 1.

51. Output growth slowed further during 1990/91-1993/94, as TFP declined by 1½ percent annually and capital growth moderated to 4½ percent, partially reflecting sluggish world demand and the imposition of a sales quota by the De Beers Central Selling Organization. During 1994/95-1996/97, an improvement in efficiency in the mining sector (annual TFP growth averaged 6½ percent) and a pickup in world demand fueled average annual output growth of 4½ percent.

Manufacturing

52. Initially, Botswana’s manufacturing sector consisted of the beef industry and a few import-substituting industries. In an effort to broaden the production base and encourage labor-intensive industries in the 1980s, the authorities embarked on a vigorous promotion of both export industries and labor-intensive manufacturing activities through incentive schemes such as the Financial Assistance Policy (FAP), the Local Preferential Scheme (LPS), and the Selebi-Phikwe Regional Development Project. The FAP provides financial support to firms in the form of grants for the promotion of labor-intensive activities in the export-or import-substituting industries. The LPS promotes the government purchase of locally produced goods, provided that the price differential with imported items does not exceed 40 percent of local content costs. These incentive schemes have recorded some success as additional investment in manufacturing has been stimulated and the export of nontraditional goods has increased since the late 1980s and early 1990s. Notable among nontraditional exports are textiles and motor vehicles. Vehicle exports, mostly to South Africa, increased from US$38 million in 1993 to an estimated US$331 in 1998. The manufacturing sector’s contribution to formal sector employment was 10 percent in 1998.

53. As a result of these efforts, which began in earnest in the late 1980s, the lackluster performance of the manufacturing sector during the first half of the 1980s was followed by average annual growth of 20 percent during the 1986/87-1989/90 period (Table II.4). This robust growth was realized through a high rate of capital accumulation (23½ percent per year), as well as through substantial employment creation (19 percent per year), despite stagnant TFP growth.

Table II.4.

Botswana: Growth of Capital, Labor, and TFP in Manufacturing Sector, 1982/83-1996/97 1/

(Average annual percentage change)

article image
Source: Fund staff estimates.

National accounts year beginning April 1.

54. The combination of high growth rates and slow technological progress in the manufacturing sector may partially reflect the effects of the government’s FAP. The incentives offered under this scheme (especially the 80 percent labor cost support for the initial five years of operations) may have encouraged manufacturing firms to overemploy labor during the eligibility period, thus increasing output at the cost of a decline in productivity. Hence, Botswana’s increased growth in manufacturing was mainly due to extensive use of factors of production. In contrast to the developments during the 1980s, employment creation in the manufacturing sector declined by almost 2 percent during the period 1990/91-1993/94 while capital accumulation slowed considerably. These developments led to a stagnation of output, even though TFP growth improved modestly during this period. The low growth rate of the manufacturing sector during this period is partially explained by the poor performance of the textile industry, which was adversely affected by trends in export markets, particularly Zimbabwe, where demand was reduced temporarily by the drought and the devaluation of the Zimbabwe dollar.33 In 1994/95-1996/97, the sector started to recover, following the diversification of the textile market and the initiation of new export products, such as assembled motor vehicles. This most recent period saw a moderate increase in employment creation and almost 2 percent annual TFP growth, while capital growth continued strong at 8½ percent per year.

Agriculture

55. The agricultural sector’s contribution to Botswana’s real GDP declined significantly from about 40 percent at independence in 1966 to 30 percent in the mid-1970s, and to only 3 percent in 1996/97. This decline reflects the enormous expansion of output in the mining and government sectors, and rapid urbanization. Despite this sharp decline, the sector still provides income and employment for the majority of the population. The agricultural sector in Botswana comprises two distinct subsectors; the traditional subsector, which practices mainly mixed subsistence farming, with individually managed arable holdings and communal grazing of livestock, and a commercial subsector, which uses modern farming techniques. The commercial subsector tends to specialize in cattle production, although cereal production is also an important activity. The main cereal crops are sorghum, maize, and millet. Other crops include groundnuts, sunflowers, and horticultural products. Successive periods of drought are frequent in Botswana. As a result, food crop production covers less than one-third of local consumption even in drought-free years, making Botswana a net importer of cereals. On the other hand, Botswana is a net exporter of beef, with beef processing accounting for over 80 percent of agricultural output, more than 95 percent of which is exported. The Botswana Meat Commission’s abattoir at Lobatse processes 800 head of cattle and 500 small ruminants per day, making it the largest in Africa.

56. The growth accounting framework seems to provide some explanation for the disappointing performance of the agricultural sector in recent years. Following a strong recovery from drought in the late 1980s, annual output declined by 1 percent during the 1990/91-1996/97 period, as labor declined by 7 percent a year (Table II.5). The contraction in output and the strong decline in employment during the 1990s is partly explained by the limited agricultural opportunities in Botswana, which, in turn, are due to the small arable land base,34 as well as by the strong urban migration bias resulting from the attraction of significantly higher manufacturing, services and public sector incomes.

Table II.5.

Botswana: Growth of Capital, Labor, and TFP in Agricultural Sector, 1982/83-1996/97 1/

(Average annual percentage change)

article image
Source: Fund staff estimates.

National accounts year beginning April 1.

57. Despite the disappointing performance in the agricultural sector in the 1990s and the continuing decline in employment, the improvement in TFP during the 1990/91-1993/94 period and the boost in capital expenditure in 1994/95-1996/97 have been encouraging. These developments likely reflect the recent movement away from traditional forms of agriculture activities toward more capital-intensive farming, and could prove to be the source for higher agricultural sector output over the medium term if the decline in labor can be mitigated.

D. Cross-Country Comparisons

58. Table II.6 below provides a cross-country comparison of Botswana’s growth and TFP performance with a number of strong-performing Asian countries, as well as with the United States and South Africa. The table highlights the strong performance of the Asian countries during the past two decades, attributable to both their relatively high TFP growth and high capital accumulation. At the same time, the low growth rate of the United States over the same period can be seen to partially reflect its modest growth in TFP. The long period of economic contraction in South Africa is also partially explained by the decline in TFP over the period, although the decline in output continued in recent years despite an improvement in TFP.

Table II.6.

Botswana: Comparison of Output and TFP Growth with Selected Countries, 1978-96 1/

(Average annual percentage change)

article image
Sources: Sarel (1997); and Fund staff estimates.

Output and TFP data for Botswana are based on national accounts years beginning July 1.

59. Botswana’s economic performance far outstripped that of the Asian countries over the past two decades, both in terms of output and TFP growth.35 More recently, however, TFP growth fell sharply in Botswana, while increasing in all of the other countries shown. Nonetheless, Botswana was able to maintain strong, though lower, GDP growth, largely on account of sizable capital accumulation. Excluding the mining sector, Botswana still achieved higher growth rates, which indicates that not only the mining sector was driving the growth process.

60. It should be noted that TFP has its limitations as a measure of productivity and should be interpreted with care, particularly as it is not directly observable and must be calculated as a residual after relevant values of ΔKKandΔLL have been estimated. It is therefore sensitive to the accuracy of the measurement of capital, labor, and a. These considerations make cross-country comparisons of TFP measures difficult to the extent that they are based on different accounting and empirical methodologies. Consequently, marginal productivity of capital (MPK) can be viewed as an alternative measure of efficiency, as it measures rates of return to investment. Hence, MPK can be useful in cross-country comparisons as capital flows tend to take advantage of the relative returns. Using the same production function, MPK can be expressed as α(Y/K).36

61. Table II.7 below provides a cross-country comparison of MPK and profit shares for the same group of countries considered in Table II.6. It suggests that Botswana’s returns on capital were particularly high during the 1978-96 period. Although the rate declined somewhat in recent years, it was still much higher than the Asian countries on average, as well as much higher than that of the United States and South Africa. Improvements in TFP should help maintain this high rate of return.

Table II.7.

Marginal Product of Capital and Profit Shares in Botswana and Selected Countries, 1978-96 1/

(In percent)

article image
Sources: Sarel (1997); and Fund staff estimates.

Data for Botswana are based on national accounts years beginning July 1.

E. The Quality of Labor, Capital, and Total Factor Productivity

62. The measurement and interpretation of TFP is difficult, as productivity improvements could be the result of increases in the quality of inputs rather than increases in the efficiency of their utilization. For this reason, Jorgenson and Griliches (1967) and Jorgenson, Gollop, and Fraumeni (1987) suggest disaggregating inputs by quality classes, so as to isolate the effect of improvements in labor or capital quality in growth accounting exercises and thereby avoid an overestimation of technological progress.

63. Taking account of these concerns by differentiating between skilled and unskilled labor, we indeed find that part of the growth process can be attributed to improvement in the skills of the labor force. The values of TFP found after making this differentiation are consistently lower than those obtained when the labor force is aggregated (Figure II.2). Between 1975/76 and 1994/95, the total labor force increased on average by 3 percent a year. Over the same period, the average growth rates of skilled and unskilled labor were 7½ percent and 1½ percent, respectively.37 This differential may be attributable to the government’s heavy investment in education, which has improved remarkably the educational attainment of the Botswana labor force. Between 1979/80 and 1997/98, education spending averaged 22 percent of the government’s recurrent outlays and 14 percent of development expenditure.

Figure II.2.
Figure II.2.

Botswana: TFP of Uniform and Differentiated Labor, 1975/76-1994/95 1/

(Annual percentage change)

Citation: IMF Staff Country Reports 1999, 132; 10.5089/9781451806342.002.A002

Sources: Botswana authorities; and Fund staff estimates.1/ National accounts year beginning July 1.

64. The calculation of TFP also does not take into account the expected profile of returns from different investments, even though computed TFP is quite sensitive to the timing of the expected returns. This issue is important in the case of Botswana and requires further analysis.38 The low level of TFP in the early 1980s may, for example, reflect the fact that significant investment in infrastructure was being made without a corresponding immediate growth in real output.

F. Conclusion

65. Botswana has succeeded in laying the basis for a prosperous economy. The impressive growth rates witnessed since independence have for the most part been made possible by significant increases in the capital stock and the growth of the labor force, as well as by substantial technological progress. The considerable capital accumulation reflects the government’s prudent use of its mineral revenues to promote productive investment and encourage economic diversification. Although not as important as the increases in factor inputs in explaining Botswana’s growth performance, TFP growth was nonetheless significant and was comparable to those of the rapidly growing Asian countries. In a cross-country comparison with a number of Asian countries, the United States and South Africa, Botswana’s TFP growth and marginal productivity of capital were shown to be much higher over the past two decades and partially explained by higher economic growth during this period. Government policies have played a prominent role in the growth process, especially in stimulating the growth of factor inputs in the manufacturing sector, and help explain fluctuations in TFP for both the economy as a whole and individual sectors. However, the effect of some policies has been ambivalent: while subsidizing labor in the manufacturing sector led to considerably higher employment, this policy may also have been an impediment to productivity in this sector. The recent improvement in TFP of the manufacturing sector is an indication that employment and productivity objectives can be achieved while sustaining growth. Although productivity in the agricultural sector has improved lately, expansion of this sector remains limited, owing to Botswana’s small natural resource base. Finally, the differentiation of different types of labor according to skills shows that the sustained growth has been partly achieved as a result of allocating substantial resources to education and health.

APPENDIX II Data Sources

Table II.8.

Botswana: Capital Stock at Constant 1985/86 Prices, 1981/82-1996/97 1/

(In millions of pula)

article image
Source: Central Statistics Office.

National accounts year beginning July 1.

Table II.9.

Botswana: Formal Sector Employment, 1981/82-1996/97 1/

(In thousands)

article image
Source: Botswana Central Statistics Office.

National accounts year beginning July 1.

Table II.10.

Botswana: Labor Force by Skill Factor, 1981/82-1996/97 1/

(In thousands)

article image
Source: Botswana authorities.

National accounts year beginning July 1.

References

  • Bank of Botswana, 1996, Annual Report 1995 (Gaborone: Bank of Botswana).

  • Barro, Robert 1998, “Notes on Growth Accounting,” NBER Working Paper No. 6654 (Cambridge, Massachusetts: National Bureau of Economic Research).

    • Search Google Scholar
    • Export Citation
  • Bingana, M., and M. Wright 1998, “The Capital Stock and Factor Productivity in Botswana” (unpublished; Gaborone: Bank of Botswana).

  • Easterly, W., and Stanley Fischer, 1994, “Soviet Economic Decline: Historical and Republican Data,” NBER Working Paper No. 4735: (Cambridge, Massachusetts: National Bureau of Economic Research).

    • Search Google Scholar
    • Export Citation
  • Jorgenson, Dale, and Zvi Griliches, 1967, “The Explanation of Productivity Change,” Review of Economic Studies, Vol. 34 (July), pp. 24983.

  • Jorgenson, Dale, Frank Gollop, and Barbara Fraumeni, 1987, Productivity and U.S. Economic Growth (Cambridge, Massachusetts: Harvard University Press).

    • Search Google Scholar
    • Export Citation
  • Sarel, Michael 1997, “Growth and Productivity in ASEAN Countries,” IMF Working Paper 97/97 (Washington: International Monetary Fund).

29

Prepared by John Matovu and Lamido Yuguda.

30

Despite the rapid economic growth, a large proportion of the population remains poor. According to the 1993/94 Household Expenditure Survey, 47 percent of the population lived below the poverty line.

31

The weaknesses of these two approaches are examined in detail in Sarel (1997).

32

The other activities include copper-nickel matte, soda ash, and coal mining.

33

Apart from external factors, other constraints inhibiting the expansion of the manufacturing sector include the shortage of serviced land and high rents, high utility and transportation costs, and low labor productivity.

34

Only 5 percent of land in Botswana is arable.

35

These high growth rates in Botswana were mainly driven by the mining sector, whose average growth rate over this period was more than 16 percent.

36

By definition, MPK is equal to the partial derivative of the production function with respect to K.

37

The differentiation of the labor force into skilled and unskilled workers is done on the basis of whether or not individuals have completed more than primary school. No such differentiation on the basis of the health status of workers is available for Botswana. If such were available it would have been possible to estimate the effect on productivity of the changing health profile of the workforce.

Botswana: Selected Issues and Statistical Appendix
Author: International Monetary Fund
  • View in gallery

    Botswana: Output and Productivity, 1982/83-1996/97 1/

    (Annual percentage change)

  • View in gallery

    Botswana: TFP of Uniform and Differentiated Labor, 1975/76-1994/95 1/

    (Annual percentage change)