Table 2 Sample characteristics of nationally representative datasets used to model demand systems

From: Nutrient adequacy for poor households in Africa would improve with higher income but not necessarily with lower food prices

 

Malawi

Niger

Uganda

Tanzania

Nigeria

Survey rounds

2010–2011, 2012–2013, 2016–2017

2011, 2014

2005–2006, 2009–2010, 2011–2012, 2013–2014, 2015–2016, 2018–2019

2008–2009, 2010–2011, 2012–2013

2010–2011, 2012–2013, 2015–2016a

Unique households

3,104

3,973

3,279

3,165

4,407

Household-year observations

8,089

13,086

14,420

9,196

25,977

Food items reported (number)

58

73

47

50

74

Food groups modelled (number)

18

19

19

19

19

Share of households in Q1 (round 1)

34.16

20.99

36.93

16.40

25.72

Share of households in Q2 (round 1)

28.31

30.74

27.27

29.01

30.82

Share of households in Q3 (round 1)

21.46

30.18

21.42

31.23

28.01

Share of households in Q4 (round 1)

16.07

18.09

14.38

23.36

15.46

  1. Quartiles (denoted Q1–Q4) divide the observations by the household’s total expenditures within each round, with Q1 representing the poorest households, which consume less than the international extreme poverty line (US$1.90 per capita per day, PPP). Q2 households represent the international poverty line (above the extreme poverty line <US$3.20 per capita per day, PPP). Q3 households are characterized by consuming above the international poverty line, namely <US$5.50 per capita per day, PPP, and Q4 households consume >US$5.50 per capita per day, PPP. In Nigeria, consumption data were collected twice within each survey rounds, so we effectively have six rounds of panel data. Household-year observations reflect the size of the pooled sample that includes all cross-sectional waves, with households observed multiple times over the survey.
Back to article page