Elsevier

Economic Modelling

Volume 92, November 2020, Pages 99-108
Economic Modelling

Food security in Kenya: Insights from a household food demand model

https://doi.org/10.1016/j.econmod.2020.07.015Get rights and content

Highlights

  • We evaluate household food security in Kenya in terms of access to food.

  • Expenditure and price elasticities are estimated using a QUAIDS model.

  • We interpret the estimates as indicators of household sensitivity to market shocks.

  • Rural households that rely on informal markets are the most severely affected by food insecurity.

  • Compensated variation ranges between 34% and 131% across food groups.

Abstract

This paper evaluates the household food security situation in Kenya in terms of access to food. We apply a quadratic almost ideal demand system (QUAIDS) model to nationally representative household survey data from Kenya, and estimate and interpret price and expenditure elasticities as indicators of household sensitivity to market shocks. Our estimation results show positive expenditure elasticities, close to unity, while all compensated and uncompensated own-price elasticities are negative and smaller in magnitude. A complementary welfare analysis shows high compensated variations in the long run, ranging between 34% and 131% across food groups. This suggests that rising relative food costs have led to deterioration of the food security situation in Kenya, and the most severely affected households seem to be those that rely on informal markets and reside in rural areas. To improve food security, targeted income support could be a more effective policy than price support, given the much higher estimated expenditure elasticities.

Introduction

More than 10 million Kenyans (approximately 25% of the country’s population) lack access to sufficient food in terms of quantity and quality, and are predominantly reliant on food aid at any given time of year (Sibhatu et al., 2015; FSIN, 2017). This inaccessibility of food is closely linked to loss of welfare and an increase in poverty incidences (WBG, 2018). The 2015/16 Kenya Integrated Household and Budget Survey (KIHBS) report demonstrates that a significant proportion of Kenyans are food insecure. The report also shows that the national food poverty headcount rate for individuals was 32%, which implies that around 14.5 million individuals were below the food poverty line.1 Food poverty incidences are highest in rural areas, and represent 64.2% of those living below the food poverty line. Similarly, estimates from the Global Report on Food Crises (FSIN, 2017) show that the number of food-insecure people in Kenya increased significantly over a-10-year period, from 1.3 million in 2007 to 2.2 million in 2017. These figures justify why food security should be a priority when it comes to public policy.

Food availability has long received greater emphasis compared with other food security dimensions, such as access, stability, and utilization (Pinstrup-Andersen, 2014). Moreover, numerous studies have focused on the supply capacity of food systems (for example Graham et al., 2007; Godfray and Garnett, 2014). Substantial efforts and resources have been spent on improving agricultural productivity and stimulating market access for smallholder producers (Khush et al., 2012; SDSN, 2013). However, less effort has been directed at investigating and attempting to remedy the challenges of demand, especially for populations vulnerable to food insecurity (the poor and rural). These challenges are associated with access and entitlement to food. Nobel laureate Amartya Sen (1981) famously wrote, “Starvation is the characteristic of some people not having enough food to eat. It is not the characteristic of there being not enough food to eat.” Therefore, our paper’s main contribution is in generating evidence from the demand side of the food system and providing policy-relevant insights into the household food security situation in Kenya and the African context overall.

Having a stable supply of and economic access to food is the most significant food security problem (Upton et al., 2016; Sibhatu et al., 2015). Concerns about inadequate food access have resulted in policies that focus on the income (expenditure) and prices that affect market demand by improving food access and food security, respectively (Yu et al., 2004; Pinstrup-Andersen, 2014). Rizov et al. (2014) and Cupak et al. (2015) recognize that there is a close link between all dimensions of food security and indicators such as food price and expenditure elasticities, which contain information on the market equilibrium of supply and demand. Income influences the distribution of expenditure (Pieters et al., 2013) and food expenditure patterns (Kearney, 2010; Rizov et al., 2014). A report by the Food and Agriculture Organization of the United Nations (FAO, 2012) shows that higher food prices lead to higher levels of undernourishment. Assessing the sensitivity of a household to changes in prices and income, while taking into account the role played by policies and household demographics, is therefore important when analysing food security.

This paper contributes to the growing literature that examines how food security relates to households’ consumption decisions and how prices, income, and demographics affect spending patterns.2 Our analysis of food consumption pattern enables us to establish a population’s food needs, along with the effects of income and prices. Furthermore, the measurement of food consumption and expenditure is a fundamental component of any analysis of welfare. However, an increase in household income does not necessarily mean that more of that income is spent on nutritious, health-enhancing food items. This is due, for example, to the persistence of lifestyle patterns, which are associated with heterogeneity in household characteristics (Alexandri et al., 2015; Bett et al., 2012; Ciaian et al., 2018; Regmi and Meade, 2013; Rischke et al., 2015). If the aim is to improve food security and promote a shift toward the consumption of more beneficial foods, such as those with higher nutritional value, controlling for household characteristics in food demand analysis can provide highly useful information that goes beyond the accessibility dimension.

In this paper, we focus on food demand in Kenya and apply a quadratic almost ideal demand system (QUAIDS) model to data from the Kenya Integrated Household Budget Survey (KIHBS), which is a nationally representative household survey. Price and income elasticity are estimated for five food groups to characterize heterogeneous households’ demand behaviour. Price and income elasticity values are important, because they can inform relevant policies that seek to improve food security. Furthermore, the (Hicksian) price elasticity is an input in our welfare analysis. Since food pricing and income-related policies have the potential to improve the population’s access to an adequate diet, our findings can directly inform government policies in developing an integrated strategy aimed at boosting both production (supply) and tackling food demand-oriented challenges associated with access to food.

The rest of the paper is organized as follows: Section 2 outlines the QUAIDS estimation framework; Section 3 describes the survey data and estimation strategy; Section 4 describes and discusses the empirical results, including welfare analysis of price changes; and Section 5 concludes.

Section snippets

Estimation framework: quadratic almost ideal demand system (QUAIDS)

Policy studies have identified that access to food is determined by the cost of food, willingness to pay, and household income (e.g., Westengen and Banik, 2016). These are the main components in the theoretical food demand function. Therefore, the demand system approach provides an effective method through which to consistently estimate demand characteristics by imposing and testing economic restrictions on individual behaviour. Several demand models have been applied in empirical studies.

Data and estimation strategy

We use data from the Kenya Integrated Household Budget Survey (KIHBS), a Kenya government funded household survey implemented by the Kenya National Bureau of Statistics (KNBS). The KIHBS is a comprehensive household survey implemented every ten years in Kenya, covering a nationally representative sample of 13,430 households in 2005/06 and 24,000 households in 2015/16. The key objectives of KIHBS are to update measures of living standards, the Consumer Price Index (CPI), and the System of

Estimation results and welfare analysis

The principal goal of the paper is to analyse the effects of income and prices on household food consumption behaviour and welfare. The analysis uses two cross-sectional datasets - 2005/06 and 2015/16. The idea behind such a comparative study is to establish if there are any major changes in demand over the ten-year period between the two survey waves. We use the price and expenditure elasticity matrix to compare the magnitudes between the two cross-sections as well as the change in household

Conclusion

Our results from demand system and welfare analyses demonstrate that income and price changes generate significant budget responses and lead to welfare changes which vary across types of household and food groups. Previous studies have observed that these changes tend to be larger for higher valued foods (such as meat and dairy) than for staple foods such as cereals (e.g., Abdulai and Aubert, 2004; Wong et al., 2015). We find that cereals and bread, dairy products, and essential condiments are

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

We thank the three anonymous referees for constructive comments and the editor for helpful guidelines. We thank the Kenya National Bureau of Statistics (KNBS) for granting access to the Kenya Integrated Household Budget Surveys data. The usual disclaimer applies.

References (68)

  • J. Greer et al.

    A methodology for measuring food poverty applied to Kenya

    J. Dev. Econ.

    (1986)
  • A.D. Jones et al.

    What are we assessing when we measure food security? A compendium and review of current metrics

    Advances in Nutrition

    (2013)
  • L. Li et al.

    Engel curves and price elasticity in urban Chinese Households

    Econ. Modell.

    (2015)
  • R. Ray

    Measuring the costs of children: an alternative approach

    J. Publ. Econ.

    (1983)
  • A. Regmi et al.

    Demand side drivers of global food security

    Global Food Security

    (2013)
  • R. Rischke et al.

    Supermarkets and food consumption patterns: the case of small towns in Kenya

    Food Pol.

    (2015)
  • L. Wong et al.

    Modelling the meat consumption patterns in Australia

    Econ. Modell.

    (2015)
  • W. Yu et al.

    Projecting world food demand using alternative demand systems

    Econ. Modell.

    (2004)
  • AGRA

    Africa Agriculture Status Report: Catalyzing Government Capacity to Drive Agricultural Transformation

    (2018)
  • A. Akbari et al.

    Welfare impacts of soaring food prices on Iranian urban households: evidence from survey data

    International Journal of Business and Development Studies

    (2013)
  • J. Banks et al.

    Quadratic Engel curves and consumer demand

    Rev. Econ. Stat.

    (1997)
  • H.K. Bett et al.

    Demand for meat in the rural and urban areas of Kenya: a focus on the indigenous chicken

    Economics Research International

    (2012)
  • R. Blundell et al.

    Estimation in large and disaggregated demand systems: an estimator for conditionally linear systems

    J. Appl. Econom.

    (1999)
  • M. Caliendo et al.

    Some Practical Guidance for the Implementation of Propensity Score Matching

    (2005)
  • B. Capéau et al.

    Prices, Local Measurement Units and Subsistence Consumption in Rural Surveys: an Econometric Approach with an Application to Ethiopia

    (1998)
  • L. Christensen et al.

    Transcendental logarithmic utility functions

    Am. Econ. Rev.

    (1975)
  • P. Ciaian et al.

    Food consumption and diet quality choices of Roma in Romania: a counterfactual analysis

    Food Security

    (2018)
  • T. Cox et al.

    Prices and quality effects in cross-sectional demand analysis

    Am. J. Agric. Econ.

    (1986)
  • A. Cupak et al.

    Food demand and consumption patterns in the new EU member states: the case of Slovakia

    Ekonomický časopis

    (2015)
  • A. Deaton

    The Analysis of Household Surveys: A Micro-econometric Approach to Development Policy

    (1997)
  • A. Deaton et al.

    An almost ideal demand system

    Am. Econ. Rev.

    (1980)
  • H.R. Dehejia et al.

    Propensity score matching methods for nonexperimental causal studies

    Rev. Econ. Stat.

    (2002)
  • Tirtha Dhar et al.

    An empirical assessment of endogeneity issues in demand analysis for differentiated products

    Am. J. Agric. Econ.

    (2003)
  • T. Diprete et al.

    Assessing bias in the estimation of causal effects: rosenbaum bounds on matching estimators and instrumental variables estimation with imperfect instruments

    Socio. Methodol.

    (2004)
  • Cited by (15)

    • Dynamic analysis of Caribbean food import demand

      2024, Journal of Agriculture and Food Research
    • How would the carbon tax on energy commodities affect consumer welfare? Evidence from China's household energy consumption system

      2022, Journal of Environmental Management
      Citation Excerpt :

      Correspondingly, this study uses QUAIDS to construct an energy demand system model in China to evaluate the impact of carbon tax on residents’ welfare. Unlike traditional two-rank demand models, such as the Linear Expenditure system (LES), the Extend Linear Expenditure System (ELES), Rotterdam model, the Transcendental Logarithmic System (TLS), and the Almost Ideal Demand System (AIDS), the QUAIDS model can reflect non-linearity of Engel curves and portrays a more relevant realistic of consumer demand (Korir et al., 2020). In addition, it captures the interdependence and substitution effects among various energy commodities when prices change (Okonkwo, 2021).

    View all citing articles on Scopus
    View full text