Gender Disparities and Economic Growth in Kenya: A Social Accounting Matrix Approach

Abstract

Realizing high economic growth and generating gainful employment present major challenges for Kenya. This paper analyzes the gendered employment outcomes of various investment options in Kenya using Social Accounting Matrix multiplier analysis. Results reveal that Kenya's agriculture sector accounts for the highest increase in employee compensation (mainly benefiting skilled labor and disproportionately benefiting men), while its manufacturing sector accounts for the largest share of job creation. Although women stand to benefit more from employment creation, most of these new jobs are informal with low wages. Kenya's gender disparities are a reflection of existing disparities in its labor market and socioeconomic structure. Therefore, policies aimed at addressing the constraints that limit women's effective participation in the Kenyan labor market, including increasing productivity and raising women's skills, are important for allowing men and women to benefit equally from employment and growth-promoting opportunities.

Keywords:

• Employment
• gender analysis
• social accounting
• JEL Code: J16

Acknowledgments

We would like to thank Michael Lahr for his useful comments on an earlier version of this piece. We are also grateful to the participants of the Inequality, Development and Growth Workshop held on May 19–20, 2008 in New York for the general comments raised on the version that was presented. We also received valuable comments during the IAFFE 2008 annual conference in Turin, and are grateful to the organizers of the conference. Lastly, we thank the three anonymous reviewers for their critical comments on earlier versions of this contribution. We are responsible for any errors and omissions

Notes

¹ Empirical evidence indicates that there may be an unemployment-growth trade-off in the long run (Robert J. Gordon 1995; Martin Zagler 2000; Patrick Toche 2001). If economic growth is driven by structural change, which could entail a shift toward capital-intensive production as the Kenyan experience demonstrates, then a social cost of unemployment is expected.

² For instance, an analysis based on the Kenyan Urban Labour Force Survey in 1986 revealed that when combining domestic chores with economic activities, women of working age worked 50.9 hours per week, compared to only 33.2 hours for men (Kabubo-Mariara 2003).

³ Conventionally, the labor sector in the SAM is captured in monetary values, but an employment satellite account presents employment data in numbers and links it with the production account.

⁴ For details, see Ellis et al. (2007).

⁵ The contrast between formal and informal sectors is no longer as straightforward as it seemed in the 1970s and 1980s, since the formal sector (particularly the private sector) also utilizes informal labor arrangements such as reliance on casual or high turnover workers. “Employment in the informal sector” and “informal employment” are concepts that are now recognized as different aspects of the normalization of employment (see Ralf Hussmanns 2005).

⁶ It is difficult to get precise estimates of the size of the informal sector, given the diversity. The numbers reported for the latter years are projections as given annually by Kenya National Bureau of Statistics. Additionally, definitions and coverage of the informal sector have changed over time.

⁷ Women constitute between 65 and 75 percent of workers in the cut-flower sector (Ellis et al. 2007).

⁸ Intrahousehold patterns are characterized by “separate purses” whereby men and women maintain a certain level of independent control over their earnings and resources while adhering to gender-specific expenditure patterns (Aspaas 1998).

⁹ Outdated labor laws have also affected women's ability to fully benefit from employment opportunities.

¹⁰ The absence of appropriate, gender-related macroeconomic analytical tools has penalized quantitative analyses of gender issues, especially for Kenya (Anushree Sinha and Haider Khan 2008). Notwithstanding the restrictive assumption of fixed prices, the incorporation of gender in a SAM-based multiplier models facilitates analyses of the impact of various macro policies by gender, thus making the effects more visible to policy-makers.

¹¹ For details on SAM construction, see Jane Kiringai, Bernadette Wanjala, James Thurlow, Nicholas Waiyaki, Clive Mutunga, Moses Njenga, Nancy Nafula, and John Mutua (2007).

¹² These are agriculture, fishing, forestry, mining and quarrying, meat and dairy processing, milling, bakery and confectionary, beverages and tobacco, other food manufactures, textiles and footwear, wood and paper, printing and publishing, petroleum, chemicals, metals and machinery, non-metallic manufactures, other manufactures, electricity and water, building and construction, trade, hotels and restaurants, transport and communication, financial services, other services, education, health, and public administration.

Note that even though tourism is an important source of growth for the Kenyan economy, the sector is not explicitly captured in national accounts. It is captured within the service sectors, which mainly include hotels and restaurants, trade, and transport and communication. Therefore, the tourism sector is similarly captured through these service sectors in the 2003 SAM for Kenya.

 A more highly disaggregated micro SAM is available with fifty activities and fifty commodities (twenty-two agriculture, eighteen industry, and ten services), three transaction costs (domestic, import, and export), twelve labor categories (by gender and skill level), twenty household categories (by region and per capita expenditure), capital (by region and formal/informal), enterprises (region and formal/informal), and three taxes (direct, commodity, and trade taxes).

¹³ The problem of data constraints in SAM frameworks is not uncommon given the detailed data requirements and the fact that relevant data has to be extracted from various available sources (for example, see Parikh and Thorbecke [1996]; Sinha and Khan [2008]).

¹⁴ The unskilled included those with no education and those with only nursery and primary education but with no vocational/professional training. The skilled included those (1) who did not have a formal education, but did have trade tests, or who had nursery and primary education, trade tests, and an ordinary diploma; (2) who had a secondary education, with/without trade tests, and an ordinary diploma; (3) who had a secondary education, a higher national diploma, and professional training, such as the Certified Public Accountants or Certified Public Secretaries; (4) who had a bachelor's degree, with/without an ordinary diploma; and (5) who were postgraduates, with/without additional professional training.

¹⁵ It was not possible to disaggregate households by gender because of complications related to the definition of gender and household head in the household survey. Eighty-four percent of the respondents were in rural areas, which could have an impact on the definition of male- or female-headed households. For instance, in some households, the male head resided in an urban area and transferred income to a rural home headed by his wife. Such households were probably termed female-headed; thus classifying households by gender could result in biased estimates that overestimated the proportion of female-headed households.

¹⁶ Additional data were obtained only when the same households could be identified.

¹⁷ This assumption makes sense for the Kenyan economy given the high unemployment rate and the high incremental capital output ratio, which implies that output can be increased through increasing production efficiency.

¹⁸ As previously stated, several assumptions hold in multiplier analysis, such as excess production capacity, which implies that increasing production in the productive sectors will not have an impact on prices.