Chapter 2 Methodology
- Marco Pani, and Mohamed El Harrak
- Published Date:
- May 2010
The impact of the composition of expenditure on the country’s development objectives has been assessed on the basis of a stylized model of the Beninese economy that focuses on the segmentation of the labor market into a rural and urban sector, and on the role of the government as a source of employment and as a supplier of capital and public services. The model, calibrated to replicate the recent economic performance and the baseline medium-term projections developed by Fund staff, has been used to compare the impact of an increase in the wage bill with that of an increase in public investment under different scenarios.
Like other small low-income countries in the region, Benin is characterized by the coexistence of a large, “traditional” rural sector with a more formal but less competitive urban sector. About 60 percent of the Beninese population lives in rural areas; agriculture (and related activities) account for about one-third of GDP, for one-fourth of exports of goods and services, and for about 70 percent of employment.2 About 50 percent of the population is more or less directly employed in the cotton sector alone, concentrated in the north of the country. In addition, “informal” activities— typically, small-scale informal trade with Nigeria—account for a large and nonquantified share of the economy. The urban economy hosts about 30 percent of the labor force;3 it is centered on the public administration and on capital-intensive state-owned and private enterprises, where employees are organized in strong and active labor unions. Around this “formal” urban employment a variety of “informal” small-scale and poorly remunerated activities take place, like providing transportation, housekeeping, and other services to the formal sector and its better-paid employees.
Reflecting this economic structure, the labor market is segmented into two distinct sectors, the rural sector, which is more competitive but offers comparatively low “wages” (also including the income of self-employed workers and small entrepreneurs), and the urban sector, where strong labor unions are capable of negotiating comparatively high wages. These wages do not clear the market, and a number of urban dwellers remain effectively unemployed (although no official unemployment statistics are available), eking out a living barely above subsistence level through occasional and precarious informal activities. Mobility across sectors is limited—most notably by the costs involved in relocating from rural to urban areas—but the large wage differential between the two sectors encourages a gradual, ongoing migration of young workers to the cities in pursuit of better-paid jobs.
The economy of migration with dual labor markets has been analyzed in a series of studies following the seminal works of Todaro (1969) and Harris and Todaro (1970) (for a brief illustration see Appendix 2). In the urban sector, wages above the competitive market-clearing level result in structural unemployment as labor supply exceeds labor demand. Despite this, workers continue to migrate, hoping to find a well-paid job in the cities, as far as the wage differential between urban and rural areas, combined with the probability of finding a job in the city, remains sufficiently large. Migration stops only when the number of the unemployed is so large that the risk of remaining unemployed reduces the potential benefit of earning a higher wage in the urban areas.
The urban and rural sectors also differ in capital intensity. The urban economy is dominated by capital-intensive activities in industry and trade (including, in Benin, energy production and the port); typical rural activities such as agriculture and informal trade are instead more labor intensive. Both private and public capital are employed in the production process in both sectors; private capital is provided by private investors (including state-owned enterprises) and consists of machinery and equipment that is earmarked to a specific firm or unit of production, while public capital includes assets such as transportation and communication infrastructure, that are supplied by the government and made publicly available. Public capital can be extensively defined to include intangible assets such as the judicial system and the rule of law as well as the “human capital” enhanced through public investment in health and education. In the rural sector, public capital includes public infrastructure specifically provided for agriculture, such as irrigation, fertilizer supply, and marketing structures.
Private capital comes from domestic and foreign investment. The model assumes that domestic investment responds to disposable income (the income received by the workforce net of taxes) and to the marginal rate of return on private capital, while foreign investment is assumed to be driven by exogenous factors.
The government plays three major roles: it supplies public capital through public investment; it provides employment and public services through the civil service (financed by the wage bill); and it raises taxes to finance these (and other) items of public expenditure.
The labor market for civil servants bears some similarities to the private urban labor market, not only because most civil servants work in urban areas (mostly in the official and administrative capitals, Porto Novo and Cotonou). Civil servants earn comparatively high salaries and bonuses, are organized in powerful labor unions, and enjoy a number of other benefits including comparatively generous social protection. The importance of the civil service in the larger labor market derives however mostly from the indirect influence of the pay level of civil servants on the wage negotiations between enterprises and labor unions. The wage level and structure of the civil service are used by urban employers and labor unions as a benchmark in their wage negotiations, and an increase in civil service wages has therefore strong repercussions on urban wages, labor demand, and unemployment.
A model incorporating these features has been calibrated to fit Benin’s actual data for the period 2003—07;4 other parameters have been set to yield results in line with the medium-term macroeconomic projections elaborated by the Fund staff on occasion of the 2010 Article IV Consultation—considered here as the “baseline” for this analysis. The model has then been used to estimate the impact of alternative scenarios associated with different changes in the composition of expenditure compared with baseline projections.
Four alternative scenarios have been considered; two involve the use of additional external grants amounting to about 2 percent of GDP to finance an increase in wages of civil servants (Scenario A) or additional public investment (Scenario B); the other two scenarios involve instead a reallocation of expenditure, from public investment to wages (Scenario C) or from wages to public investment (Scenario D). The magnitude of the changes is the same under the four scenarios.5 As mentioned in Chapter 1, the changes in the public wage bill in Scenarios A, C, and D are assumed to be produced by variations in the average levels of wages and bonuses paid to public employees, with no change in the number of civil servants.
As the focus of this study is on the composition of public expenditure, effects associated with changes in the overall level of revenue or public expenditure have not been considered, although they would have a major impact on the economy. Tax revenue has been assumed to remain, under all scenarios, at the same level projected under the baseline, with tax rates adjusting to the changes in the tax base. Public expenditure increases compared to the baseline in Scenarios A and B, but this increase is assumed to be financed entirely by external grants, with no burden for current or future domestic taxpayers. In Scenarios C and D, the overall expenditure envelope is the same as in the baseline.
The economic performance has been assessed on the basis of three indicators: real GDP growth, the rate of unemployment (as a measure of poverty), and the ratio of rural to urban wages (as a measure of disparities in income distribution).
The calibration exercise is made more complicated by the limited availability of data. Thus, particular attention was given to the choice of data and the setting of proxies for unavailable time series. Labor force data were derived from Mongardini and Samake (2009) and from the World Development Indicators database of the World Bank. The distribution of the labor force between the rural and urban sectors was estimated to be proportional to the distribution of rural and urban population. The stock of private and public capital was estimated from the series of private and public investment, drawing again from Mongardini and Samake (2009). Owing to the lack of data, the average level of real wages in the rural and urban areas, and the marginal real return on capital, were calibrated to give the best fit to the assumptions of the model.
Fund staff estimates.
Fund staff estimates.
2008 has been excluded from the estimation of some parameters owing to the particular conditions associated with the increase in international food and energy prices.
Scenarios A and B entail an increase in public expenditure by CFAF 60 million, equivalent to about 2 percent of GDP in 2010; Scenarios C and D entail an increase (decrease) in the public wage bill by CFAF 30 million and a reduction (increase) in public investment by the same amount; the net change in public expenditure is zero but the impulse produced on the rest of the economy is equal to the sum of the two changes, hence CFAF 60 million.