Nepal: Selected Issues and Statistical Appendix
Author:
International Monetary Fund
Search for other papers by International Monetary Fund in
Current site
Google Scholar Close
Publication Date:
24 Sep 2002
eISBN:
9781451829952
Language:
English
Keywords:
ISCR; CR; local government; basic salary; pay; yield; crop yield; wheat productivity; wheat crop; maize productivity; maize yield; Agricultural commodities; Agricultural credit
Download PDF (627.7 KB)

This Selected Issues paper and Statistical Appendix reviews agricultural productivity in Nepal and examines its links at the regional and aggregate levels to the amounts of available inputs such as chemical fertilizers, irrigation water, and improved seeds, as well as rainfall, rural credit, and foreign aid. The paper highlights factors that are statistically correlated with agricultural productivity and, as importantly, those that are not. The paper examines causes for the recent export slowdown. The bottom-heavy civil service structure is also described.

Abstract

This Selected Issues paper and Statistical Appendix reviews agricultural productivity in Nepal and examines its links at the regional and aggregate levels to the amounts of available inputs such as chemical fertilizers, irrigation water, and improved seeds, as well as rainfall, rural credit, and foreign aid. The paper highlights factors that are statistically correlated with agricultural productivity and, as importantly, those that are not. The paper examines causes for the recent export slowdown. The bottom-heavy civil service structure is also described.

I. The Determinants of Agricultural Productivity1

A. Introduction

1. Agriculture is the backbone of the Nepalese economy. The share of agriculture as a percent of GDP is close to 40 percent, and about 80 percent of the working population is employed in the sector. The key crops are paddy, wheat and maize, accounting for more than one third of overall agricultural production.

2. This paper analyzes the determinants of agricultural productivity in Nepal.2 In particular, the study examines the relationship between the productivity (yield) of three major crops (paddy, wheat, and maize) and agricultural inputs (chemical fertilizers, improved seeds, irrigation, agricultural credit and foreign aid disbursements, and rainfall).3

3. The main findings of the paper can be summarized as follows:

  • Weather remains an important determinant of agricultural productivity.

  • The use of chemical fertilizers has broadly contributed to the rise in paddy and wheat productivity; in contrast, maize productivity has been mostly unaffected.

  • The use of improved seeds has not been associated with higher yields of paddy and maize crops, although the yield of wheat crops has been positively affected.

  • Generally, the improved access to irrigation facilities has not been linked to higher crop productivity. Its beneficial effect is observed only in some regions.

  • Higher agricultural credit and foreign aid disbursements are associated with better crop productivity.

4. The remaining part of the paper is organized as follows. Section B describes the changes in crop productivity over time and presents some background information in the areas of fertilizer, improved seeds, irrigation and foreign aid policies. Section C introduces the data and variable definitions used in the analysis. Section D discusses the main findings of the paper. Section E concludes.

B. Background

5. The productivity of paddy, wheat, and maize crops generally increased, although slowly, during the past decades (Figure I.1). Most pronounced was the increase in paddy yieldsβ€”the mean yield between 1978/79 and 1988/89 was 1.91, i.e., over 0.50 metric tons per hectare less than the one between 1989/90 and 1999/2000. The productivity of wheat crops rose steadily in the 1990s, with mean yields during the two subperiods (1978/79–1988/89 and 1989/90–1999/2000) of 1.27 and 1.50 metric tons per hectare, respectively. The productivity of maize crops increased modestly during the sample period, as the maize yield in 1999/2000 was only 0.09 metric tons per hectare higher than in 1978/79.

Figure I.1
Figure I.1

Paddy, Wheat and Maize Yields

Citation: IMF Staff Country Reports 2002, 206; 10.5089/9781451829952.002.A001

6. Following a continuous rise in the consumption of chemical fertilizers, the fertilizer market collapsed in the mid-1990s and was deregulated in late 1997. The institutional distribution and sale of inorganic (chemical) fertilizers in Nepal started in the late 1960s.4 The consumption of chemical fertilizers increased dramatically between 1978/79 and 1994/95, but dropped significantly in the next two years (Figure I.2). This collapse of consumption, as well as problems with the distribution and the efficient use of the fertilizers, prompted the deregulation of the market in November 1997, when the Agriculture Inputs Corporation (AIC) lost its monopoly over the imports and distribution of fertilizers in Nepal. Under the new institutional setup, the Fertilizer Unit of the Ministry of Agriculture was given the mandate to oversee the distribution of chemical fertilizers and monitor their quality.

Figure I.2
Figure I.2

Use of Chemical Fertilizers

Citation: IMF Staff Country Reports 2002, 206; 10.5089/9781451829952.002.A001

7. The consumption of all types of improved seeds fluctuated during the past several decades. The supply of improved seeds in the formal sector did not start until the 1970s, with the start of improved seed imports from India and the establishment of the Hetauda Seed Processing Plant with the cooperation of the Food and Agriculture Organization of the United Nations (FAO). However, the consumption of wheat seeds, which comprises more than 80 percent of the total amount of improved seeds used each year, does not show an upward trend (Figure I.3). The consumption of maize and paddy seeds has also not increased significantly during the period from 1978/79 to 1999/2000.

Figure I.3
Figure I.3

Consumption of Improved Seeds

Citation: IMF Staff Country Reports 2002, 206; 10.5089/9781451829952.002.A001

8. The development and maintenance of irrigation facilities have proceeded for many decades, including with the support of foreign donors.5 After a period of limited progress, government investment in large-scale irrigation facilities rose between 1970 and 1975, as borrowing from international institutions increased. During this period, the focus was on the construction of new facilities, but not on their subsequent maintenance. To address the maintenance problem, a number of management-oriented projects were initiated in the 1980s, with the assistance of international organizations.6 Most ongoing irrigation projects are multi-stage initiatives, which were started decades ago. The financing of each stage often comes from different sources, although the involvement of international institutions such as the World Bank and AsDB is prevalent.

9. The agricultural sector is one of the large recipients of foreign grants and loans in Nepal. For example, approximately one fifth of the total foreign aid disbursements in recent years were designated to agriculture and irrigation.7 In particular, foreign grants and loans frequently finance the development and maintenance of irrigation facilities, as well as the purchase of improved seeds.

C. Data Source and Sample Description

10. The data used in the empirical analysis were derived from various issues of the Statistical Year Book of Nepal. The crop yields were computed using aggregate and regional data on the estimated area and production volume, reported in the yearbook. The annual series for the consumption of chemical fertilizers and improved seeds, agricultural credit and foreign aid disbursements, and annual rainfall were available in the same publication.8 The descriptive statistics of the main variables are presented in the appendix.

11. Two samplesβ€”aggregate and regionalβ€”were constructed and used in the empirical analysis. The aggregate sample contains 22 observations, covering the period from 1978/79 until 1999/2000. The regional sample includes 110 observations, with 22 data points for each of the five development regions (Eastern, Central, Western, Mid-western, and Far-western).

D. Results9

12. This section presents the estimation results, obtained from different samples and model specifications (Tables I.1–I.4). In particular, the discussion focuses on the effect of chemical fertilizers, improved seeds, irrigation, agricultural credit, foreign aid, rainfall, and regional factors on agricultural yields.

Table I.1.

The Determinants of Crop Yields

(Aggregate data)

article image
Notes: 1. Parameters estimated using ordinary least squares, 2. Robust standard errors are shown in parentheses. 3. Significance levels of 1 percent, 5 percent and 10 percent are denoted by (***), (**) and (*), respectively. 4. Agricultural credit and foreign aid are measured in billions of constant Nrs (1995/96= 100). 5. Rainfall is measured in thousands of mm. 6. Fertilizer and seed inputs are measured in thousands of metric tons. 7. Irrigation area is measured in thousands of hectares.
Table I.2.

The Determinants of Paddy Yields

(Regional data)

article image
Notes: 1. Standard errors are adjusted for clustering by region. 2. Measurement units are described in Table A.1.
Table I.3.

The Determinants of Wheat Yields

(Regional data)

article image
Notes: 1. Standard errors are adjusted for clustering by region. 2. Measurement units are described in Table A.1.
Table I.4.

The Determinants of Maize Yields

(Regional data)

article image
Notes: 1. Standard errors are adjusted for clustering by region, 2. Measurement units are described in Table A.1.

Rainfall

13. Rainfall remains an important determinant of agricultural productivity in Nepal. The importance of good monsoon rains for paddy and wheat yields is illustrated by the sign, size and significance of the variable Rainfall in Tables I.1–I.3 The magnitude of the coefficient in Table I.1, Column (3), for example, implies that an additional 100 mm of rainfall is associated with an increase in the paddy yield by 0.074 metric tons per hectare. The insignificant effect of Rainfall in the maize yield regressions is surprising, given that the bulk of maize production in Nepal is rainfed.

Chemical Fertilizers

14. The use of chemical fertilizers has broadly contributed to the rise in paddy and wheat productivity, while maize productivity has been mostly unaffected. In the paddy equations, the magnitude of the positive effect is such that a ten-thousand-ton increase in fertilizer use is associated with a productivity rise in paddy production in the range of 0.03–0.09 metric tons per hectare (Table I.1, Columns (1)–(3)).10 In the wheat yield equations, the positive effect of fertilizer use is statistically significant only in some specifications. In the maize equations, the coefficients of the variable Fertilizer are insignificant in all estimations that use aggregate data. Nevertheless, the regional sample regressions indicate that the yield of maize crops in the Central region have benefited from the increase in fertilizer use (Table I.4).

15. The established differential impact of the use of chemical fertilizers on crop productivity is not surprising. Unlike paddy and wheat only a small proportion of maize growers use fertilizers in Nepal. For example, only 18 percent of the total maize area was fertilized in 1991/92. In contrast, 46 percent of the paddy area was fertilized during the same year.11 Moreover, chemical fertilizers in maize production are likely to be more popular in some regions, but not in others. The varying effect of fertilizer use on the productivity of maize production across regions is recovered using the regional data.

Improved Seeds

16. The use of improved seeds is associated with higher yields of wheat crops only. In particular, the coefficient of Seeds suggests that a two-hundred-ton increase in the use of improved seeds is linked to a rise in wheat productivity by 0.01 metric tons per hectare (Table I.1, Column (6)). In the paddy and maize equations, however, the variable Seeds has the wrong sign in all specifications, although these coefficients are not statistically significant.12 The finding that unproved seeds increase the yield of wheat crops is consistent with the fact that the use of improved seeds is most common in growing wheat crops, with over 80 percent of all improved seeds going toward the production of wheat (Annex Table I.1).

Irrigation

17. The beneficial effect of improved access to irrigation facilities is found in several specifications only. Using the aggregate sample, the effect of irrigation is found to be statistically insignificant for all crop yields. In almost all cases, the coefficient of Irrigation is insignificant in the regional regressions as well. One of the exceptions is the positive and significant effect of irrigation on maize productivity (Table I.4, Column (1)), which implies that a one-thousand-hectare increase in irrigated land is associated with an increase in the crop yield by 0.007 metric tons per hectare. The possibility of a differential (regional) effect of irrigation on crop yields is explored by including region-irrigation interaction terms in the regressions for each crop (Tables I.2, Tables I.3 and I.4, Column (2)). The results suggest that all types of crops in the Mid-western region have benefited from the expansion of irrigation facilities (compared to those of the reference category Eastern region). In addition, the positive impact of irrigation on paddy productivity is stronger in the Far-western and Western regions.

Agricultural Credit

18. Access to agricultural credit is linked to higher agricultural yields for all crops.13 The coefficient of the credit variable is positive and statistically significant in all specifications in Table I.1. In the paddy and wheat equations, for example, the size of the coefficient on Agricultural credit implies that an additional Nr. 100 million (1995/96 = 100) is related to an increase in paddy (or wheat) productivity by 0.01 metric tons per hectare. However, one needs to be cautious in interpreting this as a convincing piece of evidence of a causal relationship. Although credit is obviously important for agricultural production, endogeneity problems could be biasing the estimated coefficients.

Foreign Aid14

19. Larger foreign aid disbursements are associated with higher crop yields. The coefficient estimates suggest that a one-billion-rupee increase in foreign aid disbursements is correlated with an increase in the paddy yield in the amount of 0.06 metric tons per hectare. This finding could be interpreted as indicating that foreign aid has been effective in raising agricultural productivity in Nepal. However, as in the case of agricultural credit, one needs to keep in mind that endogeneity problems could be biasing the coefficients on Foreign Aid

E. Conclusion

20. The findings of this paper lend support to the following conclusions. First, weather conditions continue to be a crucial factor, determining to a large extent the yields of crops in Nepal. Second, the increased use of inputβ€”chemical fertilizers, improved seeds, and irrigationβ€”has broadly contributed to the overall increase in agricultural productivity in Nepal, although this effect has not been uniform across regions and crops. Third, the access to agricultural credit and foreign aid disbursements play an important role in raising crop productivity.

21. Although drawing strong policy conclusions from the empirical analysis is difficult due to data limitation and possible endogeneity problems, the results suggest that credit availability is a key factor in raising agricultural productivity. In this context, the development of well-designed rural credit systems and their effective management in Nepal could play an important role in raising agricultural productivity and the living standards in the country.

Annex I.1 Descriptive Statistics and Model Specification

Descriptive Statistics

The summary statistics of the aggregate sample are presented in Annex Table I.1 below. The units of measurement are described in the second column. Note that the nominal series for agricultural credit and foreign aid have been deflated (1995/96 = 100).

Table I.1.

Summary Statistics

article image

Model Specification

In the aggregate sample, three different specifications for each crop yield (paddy, wheat and maize) were estimated. In the first specification (Table I.1, Columns (1), (4) and (7)), the agricultural yields are regressed on a constant and four input variables (Irrigation, Rainfall, Fertilizer and Seeds). The financial input Agricultural credit is added to the explanatory variables in the second specification (Table I.1, Column (2), (5) and (8)). The third specification uses Foreign aid as a regressor (and drops Agricultural credit), given the high degree of collinearity between the two variables (p = 0.87).

In the regional sample, five specifications were estimated for each of the three crops, using four regional dummies (with a reference category Eastern region). The first model specification includes a constant, the regional dummies, and the region-varying regressors Irrigation and Rainfall (Tables I.2, I.3 and I.4, Column (1)). In the second specification, interaction terms between the regional dummies and Irrigation are included (Tables I.1–I.4, Column (2)), in order to account for the possibility that different crop varieties (found in different regions) have different irrigation requirements. The aggregate nonfinancial variables Fertilizer and Seeds are added to the third specification (Tables I.1–I.4, Column (3)). In the forth and fifth specifications, the regressions contain Agricultural credit and Foreign aid as, explanatory variables, respectively. The standard errors of the last three specifications are adjusted for clustering (Moulton (1991)).

References

  • Economic Survey FY 2000/01 (http://dot.gov.np/News/Budget_Speech_2057–58.htm) Moulton, B., 1990, β€œAn Illustration of a Pitfall in Estimating the Effects of Aggregate Variables on Micro Unit,” Review of Economics and Statistics 72, pp. 334–338.

    • Search Google Scholar
    • Export Citation

  • National Planning Commission, Statistical Year Book of Nepal (1989, 1991, 1993, 1995, 2001), (Kathmandu: Central Bureau of Statistics).

1

Prepared by Petya Koeva (EU1).

2

Another important and interesting question, which is beyond the scope of this paper, is whether agricultural productivity in Nepal is low compared to other developing countries and (if so) what factors can explain these productivity differences. Although some data on crop yields across countries are available, answering the above question is difficult, since the crop yields may not be comparable across countries. For example, certain paddy varieties that are successfully used in other countries may not be viable in Nepal due to differences in local climate and topography.

3

The crop yield is defined as output (in metric tons) divided by area (in hectares).

4

Since the country does not produce any inorganic fertilizers, the agricultural community depends on their purchase from other countries imports, often financed by foreign grants.

5

Most of the information presented in this section comes from the website of the Department of Irrigation (http://www.doi.gov.np/doi). Unfortunately, data on the stock of irrigation facilities are unavailable. Government publications contain information on the yearly additions to the total irrigated area. Since little is known about the composition of the irrigation stock and the depreciation rates of the different facilities, the construction of capital stock series is not attempted. Instead, the empirical analysis uses the additional irrigated facilities as an explanatory variable.

6

Examples of such projects are: USAID Irrigation Management Project (1985), World Bank Irrigation Line of Credit (1988), AsDB Irrigation Sector Project (1988), UNDP/World Bank/AsDB Irrigation Sector Support Project (1989).

7

Unfortunately, the complete series for the disbursements to agriculture between 1978/79–1999/2000 are not available. The Economic Survey FY2000/01 contains this information only from 1984/85 onward. The Statistical Year Book of Nepal does not report the sectoral breakdown of foreign aid disbursements. Therefore, the empirical analysis uses the available data on total foreign aid disbursements.

8

The annual rainfall data were reported by 21 meteorological stations across the country. After identifying the location of each station, the mean precipitation for each development region was computed and used in the empirical analysis.

9

Details about the estimation and specification of the regression model can be found in the appendix.

10

Not surprisingly, the magnitude of the coefficient on Fertilizer is lower in the regressions that include Agricultural credit and Foreign aid as regressors. This can be explained by the fact that the purchase and subsequent use of chemical fertilizers are largely determined by the availability of foreign grants used to import the main inorganic nutrientsβ€”nitrogen, phosphorous, and potash.

12

The descriptive statistics of Seeds paddy, Seeds wheat and Seeds maize are reported in Annex Table I.1.

13

Unfortunately, we do not have data on credit disbursements for the production of different crops.

14

The variables Agricultural credit and Foreign aid are highly collinear, with a correlation coefficient of 0.87. Therefore, these two variables are not included together in any of the model specifications in Tables I.1–I.4.

Same Series

Other IMF Content

Other Publishers

  • Collapse
  • Expand
Cover IMF Staff Country Reports
Nepal: Selected Issues and Statistical Appendix
Author:
International Monetary Fund
Volume/Issue:
Volume 2002: Issue 206
Publisher:
International Monetary Fund
ISBN:
9781451829952
ISSN:
1934-7685
Pages:
118
DOI:
https://doi.org/10.5089/9781451829952.002
Figure I.1

Paddy, Wheat and Maize Yields
Figure I.2

Use of Chemical Fertilizers
Figure I.3

Consumption of Improved Seeds