The impact of food insecurity on health outcomes: Empirical evidence from sub-Saharan African countries

doi:10.21203/rs.3.rs-1427181/v1

Download PDF

Research Article

The impact of food insecurity on health outcomes: Empirical evidence from sub-Saharan African countries

https://doi.org/10.21203/rs.3.rs-1427181/v1

This work is licensed under a CC BY 4.0 License

Journal Publication

published 15 Feb, 2023

Read the published version in BMC Public Health →

You are reading this latest preprint version

Background: Globally, food insecurity is a leading cause of morbidity and mortality through other diseases, which means food security and nutrition are crucial to improving people's health outcomes. Both food insecurity and the health outcomes are the policy and agenda of the 2030 Sustainable Development Goals (SDGs). However, there is a lack of macro-level empirical studies concerning the relationship between food insecurity and health outcomes in sub-Saharan African (SSA) countries, even though the region is highly affected by food insecurity and its related health problems. Therefore, this study aims to examine the impact of food insecurity on life expectancy and infant mortality in SSA countries.

Methods: The study uses secondary panel data collected from international institutions. It also mainly employs Driscoll-Kraay standard errors (DKSE), a generalised method of momentum (GMM), fixed effects (FE) and Granger causality approaches, along with others estimation techniques for robustness checks.

Results: The DKSE result of model 1A confirms that a one percent increment in people’s prevalence for undernourishment reduces their life expectancy by 0.0034 percentage points. However, a one percentage increment in average dietary energy supply increases SSA countries' life expectancy by 0.0031 percentage points in model 1B. The FE results show that a one percent increment in the prevalence of undernourishment leads to an enhancement of infant mortality by 0.0119 percentage points in model 1C. However, the GMM result for model 1D confirms that a one percent increment in average dietary energy supply reduces infant mortality by 0.0004 percentage points.

Conclusions: Food insecurity harms the health status of SSA countries, but food security impacts in the reverse direction. In other words, prevalence of undernourishment adversely affects life expectancy and it increases infant mortality. Moreover, an improvement in average dietary energy supply can improve life expectancy and reduces infant mortality. The study recommends efficient utilisation of resources; improvements in investment in agricultural research, markets, infrastructures, macroeconomic policies and institutions; and developing sub-region strategies based on their agro-ecological zone, all of which are essential to overcome food insecurity and improve health outcomes.

Food insecurity

life expectancy

infant mortality

panel data estimations

SSA countries.

Food security is an essential element of people's health and well-being [1]. Further, the World Health Organization (WHO) argues that health is wealth and poor health is an integral part of poverty; governments should actively seek to preserve their people's lives and reduce the incidence of unnecessary mortality and avoidable illnesses [2]. However, lack of food is one of the factors which affects health outcomes. Concerning this, the Food Research and Action Center noted that the social determinants of health, such as poverty and food insecurity, are associated with some of the most severe and costly health problems in a nation [3].

According to the Food and Agricultural Organization (FAO), the International Fund for Agricultural Development (IFAD) and the World Food Programme (WFP), food insecurity is defined as “A situation that exists when people lack secure access to sufficient amounts of safe and nutritious food for normal growth and development and an active and healthy life” [4; p50]. It is generally believed that food security and nutrition are crucial to improving human health and development. Studies show that millions of people live in food insecurity, which is one of the main risks to human health. Around one in four people globally (1.9 billion people) were moderately or severely food insecure in 2017, and the greatest numbers were in SSA and South Asia. Around 9.2% of the world’s population were severely food insecure in 2018. Food insecurity is highest in SSA countries, where nearly one-third are defined as severely insecure [5]. Similarly, 11% (820 million) of the world’s population were undernourished in 2018, and SSA countries still share a substantial amount [5]. Even though globally the number of people affected by hunger has been decreasing since 1990, in recent years (especially since 2015) the number of people living in food insecurity has increased. It will be a huge challenge to achieve the SDGs of zero hunger by 2030 [6]. Even recently, the Global Report on Food Crises confirmed that 135 million people worldwide faced acute levels of hunger in 2019. Moreover, the report estimated that 183 million others experienced stressed food security and were at risk of falling into critical food insecurity levels, and the majority of them (around 73 million people) live in Africa [7].

An important consequence of food insecurity is that around 9 million people worldwide die every year due to hunger and hunger-related diseases. This is more than from Acquired Immunodeficiency Syndrome (AIDS), malaria and tuberculosis combined [6]. Similarly, about half of the world’s deaths occur among children in SSA due to food insecurity [8].

With the above information, researchers and policymakers should focus on the issue of food insecurity and health status. The SDGs that were developed in 2015 intend to end hunger in 2030 as one of its primary targets. However, there is a growing number of people living with hunger and food insecurity, leading to millions of deaths. In addition, despite the evidence implicating food insecurity and poor health status, there is a lack of macro-level empirical studies concerning the impact of food insecurity on people's health status in SSA countries, which leads to a knowledge (literature) gap. Therefore, this study aims to examine the impact of food insecurity on life expectancy and infant mortality in SSA countries for the period ranging from 2001–2018 using panel mean regression approaches.

Theoretical And Conceptual Framework

The conceptual framework of food insecurity and health was developed by Weiser et al. [9], who postulate that food insecurity and health are interconnected in a vicious circle via nutritional, mental health and behavioural pathways (for more detail, see [9]). Since this study focuses on the impact of food insecurity on health outcomes, and not on the causes, it adapts the conceptual framework of [9] and constructed Fig. 1.

Source: Modified and constructed by the author using Weiser et al. (9) conceptual framework.

Several findings associate the negative impact of food insecurity on health outcomes. For instance, Stuff et al. [10] found that food insecurity is related to poor self-reported health status, obesity [11], abnormal blood lipids [12], a rise in diabetes [12, 13], increased gestational diabetes [14], increased perceived stress, depression and anxiety among women [15, 16], Human Immunodeficiency Virus (HIV) acquisition risk [17–19], childhood stunting [20], poor health [21], mental health and behavioural problem [16].

This section discusses the methodology of this study. Specifically, it provides data type, sources, model specification and basic panel data econometric tests along with their justifications. It further provides estimation techniques, procedures and justifications.

Data type, sources and model specification

Except for mean years of schooling (from the United Nations Development Programme¹) and average dietary energy supply (from the FAO²), all the data were collected from the World Bank³. Notably, the data were collected from 31 SSA countries (Angola, Benin, Botswana, Burkina Faso, Cameroon, Cabo Verde, Chad, Congo Rep., Côte d'Ivoire, Ethiopia, Gabon, The Gambia, Ghana, Kenya, Lesotho, Liberia, Madagascar, Malawi, Mali, Mauritania, Mauritius, Mozambique, Namibia, Nigeria, Rwanda, Senegal, Sierra Leone, South Africa, Sudan, Tanzania and Togo). Since SSA countries suffer from food insecurity and its related health problems, this study believes the sampled countries are appropriate. Further, this study's time scope (from 2001–2018) is also appropriate because it captures the Millennium Development Goals, SDGs and other economic conditions such as the rise of SSA countries' economies and the global financial crisis of the 2000s. Therefore, this study considers various global development programmes and events.

Besides social factors, the study includes economic factors determining people's health status. Moreover, it uses two proxies’ indicators to measure both food insecurity and health status; hence, it specifies the general model as follows:

$$\stackrel{Target variables}{\underset{LNLEXP\left(LNINFMOR\right)}{⏞}}=f(\stackrel{Target variables}{\underset{PRUND\left(AVDES\right)}{⏞}},GDPPC, GOVEXP, MNSCHOOL, URBAN) \left(1\right)$$

The study uses four models to analyze the impact of food insecurity on health outcomes.

$${LNLEXP}_{it }= {\alpha }_{0}+ {\alpha }_{1}{PRUND}_{it}+{\alpha }_{2} {{GDPPC}_{it}+{\alpha }_{3}{GOVEXP}_{it}+{\alpha }_{4}{MNSCHOOL}_{it}+{\alpha }_{5}{URBAN}_{it}+ ղ}_{it} \left(1A\right)$$

$${LNLEXP}_{it }= {\beta }_{0}+ {\beta }_{1}{AVDES}_{it}+{\beta }_{2} {{GDPPC}_{it}+{\beta }_{3}{GOVEXP}_{it}+{\beta }_{4}{MNSCHOOL}_{it}+{\beta }_{5}{URBAN}_{it}+ v}_{it} \left(1B\right)$$

$${LNINFMOR}_{it }= {\theta }_{0}+ {\theta }_{1}{PRUND}_{it}+{\theta }_{2} {{GDPPC}_{it}+{\theta }_{3}{GOVEXP}_{it}+{\theta }_{4}{MNSCHOOL}_{it}+{\theta }_{5}{URBAN}_{it}+ \epsilon }_{it} \left(1C\right)$$

$${LNINFMOR}_{it }= {\delta }_{0}+ {\delta }_{1}{AVDES}_{it}+{\delta }_{2} {{GDPPC}_{it}+{\delta }_{3}{GOVEXP}_{it}+{\delta }_{4}{MNSCHOOL}_{it}+{\delta }_{5}{URBAN}_{it}+ \mu }_{it} \left(1D\right)$$

where LNLEXP and LNINFMOR refer to the natural logarithm of life expectancy at birth and infant mortality used as proxy variables for health outcomes. Similarly, PRUND and AVDES are the prevalence of undernourishment and average dietary energy supply adequacy – proxy variables for food insecurity. GDPPC is GDP per capita, GOVEXP refers to domestic general government health expenditure, MNSCHOOL is mean years of schooling and URBAN refers to urbanisation. Further, ${ղ}_{it}, {v}_{it}, {\epsilon }_{it},{ and \mu }_{it}$are the stochastic error terms at period t. The parameters ${\alpha }_{0,} {\beta }_{0}, {\theta }_{0},{\delta }_{0}$ refer to intercept terms and ${\alpha }_{1}-{\alpha }_{5}, {\beta }_{1}-{\beta }_{5}, { \theta }_{1}-{\theta }_{5}, and {\delta }_{1}-{\delta }_{5}$ are the long-run estimation coefficients. Since health outcomes and food insecurity have two indicators used as proxy variables, this study estimates different alternative models and robustness checks of the main results.

Basic panel econometric tests and their justifications

Cross-sectional dependence (CD)

A growing body of the panel data literature concludes that panel data models are likely to exhibit substantial CD in the errors resulting from frequent shocks, unobserved components, spatial dependence and idiosyncratic pairwise dependence. Even though the impact of CD in estimation depends on several factors, relative to the static model, the effect of CD in dynamic panel estimators is more severe [22]. Moreover, Pesaran [23] notes that occurrences such as recessions and economic or financial crises potentially affect all countries, even though it might start from just one or two countries. These occurrences inevitably introduce some cross-sectional interdependencies across the cross-sectional unit, their regressors and the error terms. Hence, overlooking the CD in panel data leads to biased estimates and spurious results [22, 24]. Further, the CD test determines the type of panel unit root and cointegration tests that we should apply. Therefore, examining the CD is vital and is the first step in panel data econometrics.

In the literature, there are several tests for CD, such as the Breusch and Pagan [25] Lagrange multiplier (LM) test, Pesaran [26] scaled LM test, Pesaran [26] CD test and Baltagi et al. [27] bias-corrected scaled LM test (for more detail, see Tugcu and Tiwari [28]). Besides, Friedman [29] and Frees [30, 31] also other types of CD tests (for more detail, see De Hoyos and Sarafidis [22]). This study basically employs Frees [30] and Pesaran [26] among the existing CD tests. This is because, unlike the Breusch and Pangan [25] test, these tests do not require infinite T and fixed N, and are rather applicable for both a large N and T. Additionally, Free’s CD test can overcome the irregular signs associated with correlation. However, it also employs Friedman [29] CD for mixed results of the above tests.

Unit root test

The panel unit root and panel cointegration tests are also the common steps following the CD test. Generally, there are two types of panel unit root test: (1) the first-generation panel unit root tests, such as Im et al. [32], Maddala and Wu [33], Choi [34], Levin et al. [35], Breitung [36] and Hadri [37], and (2) the second-generation panel unit root tests, such as [24, 38–47].

The first-generation panel unit root tests have been criticised because they assume cross-sectional independence [48–51]. This hypothesis is somewhat restrictive and unrealistic, as macroeconomic time series exhibit significant cross-sectional correlation among countries in a panel [50], and co-movements of economies are often observed in the majority of macroeconomic applications of unit root tests [49]. The cross-sectional correlation of errors in panel data applications in economics is likely to be the rule rather than the exception [51]. Moreover, applying first-generation unit root tests under CD models can generate substantial size distortions [48], resulting in the null hypothesis of nonstationary being quickly rejected [24, 52]. As a result, second-generation panel unit root tests have been proposed to take CD into account. Therefore, among the existing second-generation tests, this study employs Pesaran’s [24] cross-sectionally augmented panel unit root test (CIPS) unit root test for models 1A–1C. The rationale for this is that, unlike other unit root tests that allow CD, such as Bai and Ng [38], Moon and Perron [45] and Phillips and Sul [42], Pesaran’s [24] test is simple and clear. Besides, Pesaran [24] is robust when time-series’ heteroscedasticity is observed in the unobserved common factor [53]. Even though theoretically Moon and Perron [45], Choi [54] and Pesaran [24] require large N and T, Pesaran [24] is uniquely robust in small sample sizes [55]. Therefore, this study employs the CIPS test to take into account CD, heteroskedasticity in the unobserved common factor and both large and small sample countries. However, since there is no CD in model 1D, this study employs the first-generation unit root tests called Levin, Lin, and Chu (LLC), Im, Pesaran, Shin (IPS) and Fisher augmented Dickey–Fuller (ADF) for model 1D.

Cointegration test

The most common panel cointegartion tests when there is CD are Westerlund [56], Westerlund and Edgerton [57], Westerlund and Edgerton [58], Groen and Kleibergen [59], Westerlund’s [60] Durbin-Hausman test, Gengenbach et al. [61] and Banerjee and Carrion-i-Silvestre [62]. However, except for a few, most of them are not coded in Statistical Software (STATA) or Econometrics Views (EViews) and are affected by insufficient observations. The current study primarily uses Westerlund [56] and Banerjee and Carrion-i-Silvestre [62] for models 1A–1C. However, to decide uncertain results, it also uses McCoskey and Kao [63] cointegration tests for model 1C. The rationale for using Westerlund’s [56] cointegration test is that most panel cointegration has failed to reject the null hypothesis of no cointegration due to failure of common-factor restriction [64]. However, Westerlund [56] does not require any common factor restriction [65] and allows for a large degree of heterogeneity (e.g. individual-specific short-run dynamics, intercepts, linear trends and slope parameters) [50, 65, 66]. Besides, its command is coded and easily available in STATA. However, it suffers from insufficient observations, especially when the number of independent variables increases. The present study employs the Banerjee and Carrion-i-Silvestre [62] and McCoskey and Kao [63] cointegration tests to overcome this limitation. The three Engle-Granger-based cointegration tests applicable when there is no CD and which are widely used and available in EViews and STATA are Pedroni [67, 68], Kao [69] and Fisher-type [34]. Compared to the Fisher-type, both the Pedroni and Kao cointegration tests are more efficient and comprehensive. This study therefore uses them for model 1D.

Estimation techniques, procedures and justifications

This study mainly employs the Driscoll-Kraay [70] standard error (DKSE) (for models 1A and 1B), FE (for model 1C) and two-step GMM (for model 1D) estimation techniques to examine the impact of food insecurity on health outcomes. It also employs the Granger causality test. However, for robustness checks, it employs fully modified ordinary least square (FMOLS), dynamic OLS (DOLS), panel-corrected standard error (PCSE) and feasible generalised least squares (FGLS) methods (for models 1A and 1B), random effect (RE) techniques for model 1C and panel dynamic fixed effect (DFE) techniques for model 1D.

Even though several panel estimation techniques allow CD, most of them – such as cross-section augmented autoregressive distributed lag (CS-ARDL), cross-section augmented distributed lag (CS-DL), common correlated effects pooled (CCEP) and common correlated effects mean group (CCEMG) estimators – require a large number of observations over groups and time periods. Similarly, the continuously updated full modified (CUP-FM) and continuously updated bias-corrected (CUP-BC) estimators are not coded in both STATA and EViews. Others, like the PCSE, FGLS and seemingly unrelated regression (SUR), are feasible for T (the number of time series) > N (the number of cross-sectional units) [71, 72]. However, a DKSE estimate is feasible for N > T [71]. Therefore, depending on the CD, cointegration test, availability in STATA and EViews and comparing N against T, this study mainly employs the DKSE regression for models 1A and 1B. Due to the absence of cointegration, and to deal with heterogeneity and spatial dependence in the dynamic panel, this study employs FE for model 1C. However, due to the absence of CD, the existence of cointegration and N > T, and because all variables are I(1), this study uses GMM for model 1D.

The DKSE regression can be estimated in three different ways: FE with DKSE, RE with DKSE and pooled Ordinary Least Squares/Weighted Least Squares (pooled OLS/WLS) regression with DKSE. Hence, we have to choose the most efficient model using Hausman and Breusch-Pagan LM for RE tests. In other words, we have to select the most efficient model among FE, RE and Pooled OLS for models 1A and 1B; a more efficient model between FE and RE for model 1C; and the most efficient of the panel ARDL (Pooled Mean Group (PMG), Mean Group (MG), and DFE) models. Therefore, this study uses a FE model within the DKSE estimates for models 1A and 1B, and a FE model for model 1C (in the interest of space, these results are not reported here but are available from the author).

Finally, to check the robustness of the main result, this study employs FMOLS, DOLS, FGLS and PCSE estimation techniques for models 1A and 1B. Even though the Hausman test confirms that the FE is more efficient, the study employs the RE model for model 1C. This is because Firebaugh et al. [73] note that the RE and FE models are the best performers in panel data. Besides, unlike FE, RE assumes that the differences between individuals are random. This study also uses panel DFE for model 1D (selected based on Hausman test). Moreover, the robustness check is also conducted using an alternative model (i.e. dependent variable without natural log and Granger causality test).

Basic panel econometric tests

Each of Tables 1–3 provide a basic econometric test which is a precondition for panel estimations. Specifically, it has basic information about the CD, unit root (stationary) behaviour of all the variables as well as cointegration (long-run relationship) among the variables in the models.

Table 1

Cross-sectional dependence tests
tests	Model 1A	Model 1B	Model 1C	Model 1D
Pesaran's test of CD	8.862***	5.971***	3.673***	0.148
Frees' test of cross sectional independence	10.067***	10.309***	8.010***	8.386***
Friedman's test of cross sectional independence	-	-	-	16.254
CD Cross-Sectional Dependence
*** ⇒ significant at 1% level.
Source: Computed by the author using STATA 15

Table 2

Unit root tests
Pesaran [26] unit root test (Models 1A–1C)
Variables	CIPS (intercepts only)				Critical values
	Levels		1st difference		Critical values
	Statistic		Statistic		10%		5%	1%
LNLEXP	-3.839***		-2.854***		-2.03		-2.11	-2.25
LNINFMOR	-1.884		-2.548***
PRUDN	-1.528		-2.280***
AVRDES	-2.110**		-2.518***
GDPPC	-0.978		-2.800***
GOVEXP	-1.550		-3.925***
MNSCHOOL	-2.036*		-4.070***
URBAN	-1.997		-2.976***
Unit root test^E (Model 1D)
LNINFMOR		Statistic		Values		Order of integration
LNINFMOR		LLC IPS ADF		-7.69760* -5.51904* 149.259***		All variables are I(1)
AVRDES		LLC IPS ADF		-2.37413* -4.67357* 120.522***
GDPPC		LLC IPS ADF		-14.8520* -10.1380* 228.426***
GOVEXP		LLC IPS ADF		-18.0049* -16.3343* 335.915***
MNSCHOOL		LLC IPS ADF		-11.8124* -12.9479* 279.960***
URBAN		LLC IPS ADF		-21.3499* -15.3157* 209.195***
ADF Augmented Dickey–Fuller, AVRDES Average Dietary Energy Supply, CIPS Cross-Sectionally Augmented Panel Unit Root Test, GDPPC Gross Domestic product (GDP) per capita, GOVEXP Domestic General Government Health Expenditure, I(1) Integration at First Difference, IPS Im, Pesaran, Shin, LLC Levin, Lin, and Chu, LNINFMOR Natural Logarithm of Infant Mortality Rate, LNLEXP Natural Logarithm of Life Expectancy at Birth, MNSCHOOL Mean Years of Schooling, PRUDN Prevalence of Undernourishment, URBAN Urbanisation.
*, , * ⇒ significant at 1, 5, and 10% level, respectively.

Source: Computed by the author using STATA 15 and EViews 10^E

Following unit root tests, Table 3 also reports a basic econometric test called the panel cointegration test, which describe the long-run relationship among the variables.

Table 3

Panel cointegration test
Models	Values			Westerlund [56] test for only target variables							Banerjee and Carrion-i-Silvestre [62] –for all variables in the model
Models	Values			Gt		Ga		Pt		Pa	Levels Statistic
Model 1A	Z-value			-34.004***		-1.463*		-74.878***		-34.439***	-4.062***
Model 1A	Bootstrap P-value			0.000***		0.000***		0.000***		0.000***	-4.062***
Model 1B	Z-value			-34.708***		-4.303***		-52.636***		-37.899***	-4.212***
Model 1B	Bootstrap P-value			0.000***		0.000***		0.000***		0.000***	-4.212***
Model 1C	Z-value			6.958		7.497		3.508		4.582	-2.907***
Model 1C	Bootstrap P-value			0.990		0.990		0.440		0.380	-2.907***
Further cointegration tests (Model 1C)
DOLS residuals		Included variables			Unit root test		Statistic					p-value
		between target variables			IPS		t-bar			-1.5321		0.9509
							t-tilde-bar			-1.1501
							Z-t-tilde-bar			1.6541
		All variables in the model			IPS		t-bar			-1.2116		0.9984
							t-tilde-bar			-0.9728
							Z-t-tilde-bar			2.9494
Pedroni [66] cointegrtaion test^E (Model 1D)
			Test						Statistics		Weighted Statistic
Within-dimension			Panel		v-Statistic				-0.238		0.903
					rho-Statistic				4.791		4.046
					PP-Statistic				1.140		-1.568*
					ADF-Statistic				-2.103**		-2.733***
Between-dimension			Group		rho-Statistic				6.215
					PP-Statistic				-4.147***
					ADF Statistic				-3.553***
Kao [69] co-integration test^E (Model 1D)
Tests							t-Stat.
ADF							-2.637***
ADF Augmented Dickey–Fuller, IPS Im, Pesaran and Shin.
, , ** ⇒ significant at 10%, 5%, and 1% level, respectively.

Source: Computed by the author using STATA 15 and EViews 10^E

Long-run estimation results

Table 4 provides regression results of all models employing appropriate estimation techniques such as DKSE, FE, two-step GMM and Granger causality.

Table 4

DKSE, FE, Two-step GMM, and Granger causality results…main result
Variables	DKSE			FE	GMM
	Dependent variable (LNLEXP)			Dependent variable (LNINFMOR)
	Model 1A	Model 1B		Model 1C	Model 1D
PRUDN	-0.0034***	----		0.0119***	----
L_LNINFMOR	----	----		----	0.9698***
AVRDES	----	0.0031***		----	-0.0004***
GDPPC	-4.28e-06	-3.81e-06		0.000056***	1.97e-06
GOVEXP	0.0047	0.0018		-0.0103	-0.0026***
MNSCHOOL	0.0936***	0.0926***		-0.281***	-0.0034
URBAN	-0.0001	-0.0001		0.00061	0.00003
CONSTANT	3.7019***	3.2941***		4.904***	----
Dumitrescu and Hurlin [74] Granger causality test^E
Null hypothesis			W-Stat.(Zbar-Stat.)
$\text{P}\text{R}\text{U}\text{D}\text{N}\nrightarrow \text{L}\text{N}\text{L}\text{E}\text{X}\text{P}$			17.532 (48.436)***
$\text{A}\text{V}\text{R}\text{D}\text{E}\text{S}\nrightarrow \text{L}\text{N}\text{L}\text{E}\text{X}\text{P}$			13.272(35.829)***
$\text{P}\text{R}\text{U}\text{D}\text{N}\nrightarrow \text{L}\text{N}\text{I}\text{N}\text{F}\text{M}\text{O}\text{R}$			5.007(11.368)***
$\text{A}\text{V}\text{R}\text{D}\text{E}\text{S}\nrightarrow \text{L}\text{N}\text{I}\text{N}\text{F}\text{M}\text{O}\text{R}$			5.277(12.166)***
AVRDES Average Dietary Energy Supply, DKSE Driscoll-Kraay Standard Errors, FE Fixed Effect, GDPPC Gross Domestic product (GDP) per capita, GMM Generalised Method of Momentum, GOVEXP Domestic General Government Health Expenditure, L_LNINFMOR Lag of Natural Logarithm of Infant Mortality Rate, LNINFMOR Natural Logarithm of Infant Mortality Rate, LNLEXP Natural Logarithm of Life Expectancy at Birth, MNSCHOOL Mean Years of Schooling, PRUDN Prevalence of Undernourishment, URBAN Urbanisation
*** ⇒ significant at 1% level.

Source: Computed by the author using STATA 15 and EViews 10^E

Robustness checks

Since the above result is not enough for concrete conclusions and policy recommendations, Tables 5 and 6 provide estimation results to check the robustness of Table 4 results. Except for differences in the dependent variables or estimation techniques, the robustness results are similar to Table 4 concerning the relationship between target variables.

Table 5

DKSE, FE, Two-step GMM, and Granger causality results
Variables	DKSE			FE	GMM
	Dependent variable (LEXP)			Dependent variable (INFMOR)
	Model 1A	Model 1B		Model 1C	Model 1D
PRUDN	-0.1924***	---		0.9785***	----
L_INFMOR	----	---		----	0.9576***
AVRDES	----	0.1762***		----	-0.0119
Dumitrescu and Hurlin [74] Granger causality test^E
Null hypothesis			W-Stat.(Zbar-Stat.)
$\text{P}\text{R}\text{U}\text{D}\text{N}\nrightarrow \text{L}\text{E}\text{X}\text{P}$			18.330(50.799)***
$\text{A}\text{V}\text{R}\text{D}\text{E}\text{S}\nrightarrow \text{L}\text{E}\text{X}\text{P}$			14.289(38.838)***
$\text{P}\text{R}\text{U}\text{D}\text{N}\nrightarrow \text{I}\text{N}\text{F}\text{M}\text{O}\text{R}$			5.491(12.801)***
$\text{A}\text{V}\text{R}\text{D}\text{E}\text{S}\nrightarrow \text{I}\text{N}\text{F}\text{M}\text{O}\text{R}$			5.755(13.580)***
AVRDES Average Dietary Energy Supply, DKSE Driscoll-Kraay Standard Errors, FE Fixed Effect, GMM Generalised Method of Momentum, INFMOR Infant Mortality Rate, L_INFMOR Lag of Infant Mortality Rate, LEXP Life Expectancy at Birth, PRUDN Prevalence of Undernourishment.
*** ⇒ significant at 1% level.

Source: Computed by the author using STATA 15 and EViews 10^E

Table 6

FMOLS, DOLS, PCSE, FGLS, RE, and DFE results
Variables	FMOLS^E		DOLS^E			PCSE			FGLS
	Dependent variable (LNLEXP)
	Model 1A	Model 1B	Model 1A		Model 1B	Model 1A		Model 1B	Model 1A	Model 1B
PRUDN	-0.0033***	----	-0.0034 ***		----	-0.0024***		----	-0.0023***	----
AVRDES	----	0.0030***	----		0.0031***	----		0.0013***	----	0.0013***
Dependent variable (LNINFMOR)
				RE for Model 1C			DFE for Model 1D
PRUDN				0.01204***			----
AVRDES				---			-0.0177***
ECM				---			-0.0352***
AVRDES Average Dietary Energy Supply, DFE Dynamic Fixed Effect, DOLS Dynamic Ordinary Least Square, ECM Error Correction Model, FGLS Feasible Generalised Least Squares, FMOLS Fully Modified Ordinary Least Square, LNINFMOR Natural Logarithm of Infant Mortality Rate, LNLEXP Natural Logarithm of Life Expectancy at Birth, PCSE Panel-Corrected Standard Error, PRUDN Prevalence of Undernourishment, RE Random Effect,
*** ⇒ significant at 1% level.

Source: Computed by the author using STATA 15 and EViews 10^E

Cross-sectional dependence test

Except for model 1D, the results strongly reject the null hypothesis of cross-sectional independence in all models. However, for model 1D, the study found mixed results (i.e. Pesaran [26] fails to reject the null hypothesis of no CD while Frees [30] strongly rejects it). Thus, in order to decide, this study employs the Friedman [29] CD test. The result fails to reject the null hypothesis of cross-sectional independence, implying that two out of three tests fail to reject the null hypothesis of cross-sectional independence in model 1D. Therefore, unlike others, there is no CD in model 1D (see Table 1).

Unit root tests

The result shows that all variables are highly (at 1% level) significant at the first difference (I(1)) in all models, which implies all variables are stationary at I(1). In other words, the result fails to reject the null hypothesis of unit root (non-stationary) for all variables at a 1% level of significance at the first difference. Since all the variables are highly statistically significant at the first difference, we notice that all measures are I(1). Thus, we might expect a long-run connection between these variables collectively (see Table 2).

Cointegration tests

The results in Table 3 show that both the Westerlund [56] and Banerjee and Carrion-i-Silvestre [62] cointegration tests strongly reject the null hypothesis of no-cointegration in models 1A and 1B. However, model 1C provides a mixed result, i.e. the Banerjee and Carrion-i-Silvestre [62] test rejects the null hypothesis of no cointegartion, yet the reverse is true for the Westerlund [56] test. Therefore, this study conducted further cointegration tests for model 1C. Even though Westerlund and Edgerton [57] suffer from insufficient observation, it is based on the McCoskey and Kao [63] LM test [75]. Thus, we can use a residual-based cointegration test in the heterogeneous panel framework proposed by McCoskey and Kao [63]. However, an efficient estimation technique of cointegrated variables is required, and hence the FMOLS and DOLS estimators are recommended. The residuals derived from the FMOLS and DOLS will be tested for stationarity with the null hypothesis of no cointegration amongst the regressors. Since the McCoskey and Kao [63] test involves averaging the individual LM statistics across the cross-sections, for testing the residuals FMOLS and DOLS stationarity, McCoskey and Kao [63] test is in the spirit of IPS (Im et al. [32]) [76].

Though FMOLS and DOLS are recommended for the residuals cointegration test, DOLS is better than FMOLS (for more detail, see Kao and Chiang [77]); therefore, this study uses a residual test derived from DOLS. The result fails to reject the null hypothesis of no cointegartion. Two (Banerjee and Carrion-i-Silvestre [62] and McCoskey and Kao [63]) out of three tests fail to reject the null hypothesis of no cointegartion; hence, we can conclude that there is no long-run relationship among the variables in model 1C.

Unlike other models, since there is CD in model 1D, this study employs the Pedroni [68] and Kao [69] cointegration tests for model 1D. The result strongly rejects the null hypothesis of no cointegartion, implying – similar to models 1A and 1B – that a long-run relationship among the variables in model 1D exists (see Table 3).

Long-run estimation results

Table 4 presents the estimated results, however die to limitations of space, this section discusses only the results of target variables. The DKSE result of model 1A confirms that the rise in the prevalence of undernourishment significantly reduces life expectancy of SSA countries. In other words, a one percent increment in the prevalence for undernourishment reduces life expectancy by 0.0034 percentage points. However, the FE results show that an increment in the prevalence of undernourishment has a positive and significant impact on infant mortality in model 1C. The DKSE estimations result in model 1B reveals that a one percentage increment in average dietary energy supply increases SSA countries' life expectancy by 0.0031 percentage points. However, the GMM result for model 1D confirms that a one percent increment in average dietary energy supply reduces infant mortality by 0.0004 percentage points. Moreover, this study conducted a panel Granger causality test to confirm whether or not food insecurity has a potential causality to health outcomes. The result demonstrates that the null hypothesis of change in both prevalence of undernourishment and average dietary energy supply does not homogeneously cause health outcomes is rejected at 1% level of significance, implying a change in food insecurity does Granger-cause health outcomes of SSA countries (see Table 4).

Based on the above results, we can conclude that food insecurity harms SSA nations' health outcomes. This is because prevalence of undernourishment leads to increased infant mortality by reducing the vulnerability, severity and duration of infectious diseases such as diarrhoea, pneumonia, malaria and measles. Similarly, prevalence of undernourishment can reduce life expectancy by increasing the vulnerability, severity, and duration of infectious diseases. However, food security improves health outcomes – the rise in average dietary energy supply reduces infant mortality and it increases the life expectancy of individuals.

The author believes that the above findings and conclusion may not be enough for policy recommendations unless robustness checks are undertaken. Hence, the study estimated all models without the natural logarithm of the dependent variables (see Table 5, though in the interest of space, only results of target variables are reported). The model 1A result reveals that similar to the above results, prevalence of undernourishment significantly reduces individuals’ life expectancy in SSA countries. That means a one percent increment in prevalence of undernourishment reduces the life expectancy by 0.1924 years. Moreover, in model 1B, improvement in average dietary energy supply positively and significantly affects life expectancy. In model 1C, an increment in prevalence of undernourishment has a positive and significant effect on the infant mortality rate of SSA countries. The FE result implies that a one percent increment in prevalence of undernourishment increases the infant mortality rate by 0.9785. The GMM result in model 1D indicates that improvement in average dietary energy supply significantly reduces infant mortality. Further, the Granger causality result confirms that the null hypothesis of change in prevalence of undernourishment and average dietary energy supply does not homogeneously cause health outcomes is rejected at 1% level of significance, implying a change in food insecurity does Granger-cause health outcomes of SSA countries (see Table 5).

The study also conducted further robustness checks using the same dependent variables (as Table 4) but different estimation techniques. The results confirm that prevalence of undernourishment has a negative and significant effect on life expectancy, but improvement in average dietary energy supply significantly increases life expectancy in SSA countries. However, prevalence of undernourishment has a positive and significant effect on infant mortality, though improvement in average dietary energy supply significantly reduces (see Table 6).

Millions of people are dying every year due to hunger and hunger-related diseases worldwide, especially in SSA countries. Currently, the link between food insecurity and health status is on researchers' and policymakers' agendas. However, macro-level findings in this area for most concerned countries have been given only limited attention. Therefore, this study examined the impact of food insecurity on life expectancy and infant mortality rate. The study mainly employs DKSE, FE, two-step GMM and Granger causality approaches, along with other estimation techniques for robustness checks for the years between 2001 and 2018. The result confirms that food insecurity harms SSA nations’ health outcomes, while food security improves the health status. That means that a rise in undernourishment increases the infant mortality rate and reduces life expectancy. However, an increment in average dietary energy supply reduces infant mortality and increases life expectancy. Therefore, SSA countries need to guarantee their food accessibility both in quality and quantity, which improves health status. Both development experts and political leaders agree that Africa has the potential for agricultural outputs, can feed the continent and improve socio-economic growth. Besides, more than half of the world's unused arable land is found in Africa. Therefore, effective utilisation of natural resources is essential to achieve food security. Further, improvement in investment in agricultural research; improvement in markets, infrastructures, and institutions; good macroeconomic policies and political stability; and developing sub-region strategies based on their agro-ecological zone is crucial to overcome food insecurity and improve health status.

ADF

Augmented Dickey–Fuller

AIDS

Acquired Immunodeficiency Syndrome

AVRDES

Average Dietary Energy Supply

CCEMG

Common Correlated Effects Mean Group

CCEP

Common Correlated Effects Pooled

Cross-Sectional Dependence

CIPS

Cross-Sectionally Augmented Panel Unit Root Test

CS-ARDL

Cross-Section Augmented Autoregressive Distributed Lag

CS-DL

Cross-Section Augmented Distributed Lag

CUP-BC

Continuously Updated Bias-Corrected

CUP-FM

Continuously Updated Full Modified

DFE

Dynamic Fixed Effect

DKSE

Driscoll-Kraay Standard Errors

DOLS

Dynamic Ordinary Least Square

ECM

Error Correction Model

EViews

Econometrics Views

FAO

Food and Agricultural Organization

Fixed Effect

FGLS

Feasible Generalised Least Squares

FMOLS

Fully Modified Ordinary Least Square

GDPPC

Gross Domestic product (GDP) per capita

GMM

Generalised Method of Momentum

GOVEXP

Domestic General Government Health Expenditure

HIV

Human Immunodeficiency Virus

I(1)

Integration at First Difference. IFAD:International Fund for Agricultural Development

INFMOR

Infant Mortality Rate

IPS

Im, Pesaran, Shin

L_INFMOR

Lag of Infant Mortality Rate

L_LNINFMOR

Lag of Natural Logarithm of Infant Mortality Rate

LEXP

Life Expectancy at Birth

LLC

Levin, Lin, and Chu

Lagrange Multiplier

LNINFMOR

Natural Logarithm of Infant Mortality Rate

LNLEXP

Natural Logarithm of Life Expectancy at Birth

Mean Group

MNSCHOOL

Mean Years of Schooling

OLS

Ordinary Least Squares

PCSE

Panel-Corrected Standard Error

PMG

Pooled Mean Group

PRUND

Prevalence of Undernourishment

Random Effect

SDGs

Sustainable Development Goals

SSA

Sub-Saharan African

STATA

Statistical Software

SUR

Seemingly Unrelated Regression

URBAN

Urbanisation

WFP

World Food Programme, WHO:World Health Organization

WLS

Weighted Least Squares

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The author declares that there are no competing interests.

Funding

This study did not receive any specific grant.

Author contributions

SDB was collected, analysed and interpreted the data, and wrote and approved the paper for submission.

Not applicable.

Author Information

College of Business and Economics, Department of Economics, Arsi University, Ethiopia.

Availability of data and materials

The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.

Notes:

Data sources:

http://hdr.undp.org/en/data
https://knoema.com/FAOFSD2020/fao-food-security-data?location=1000180-sub-saharan-africa
https://databank.worldbank.org/source/world-development-indicators#

Giller KE. The Food Security Conundrum of sub-Saharan Africa. Global Food Security. 2020;26(2020):100431.
WHO. Reducing Risks, Promoting Healthy Life. World Health Report. Switzerland: WHO; 2002. Available from: https://apps.who.int/iris/bitstream/handle/10665/42510/WHR_2002.pdf?sequence=1.
Food Research and Action Center. Hunger and Health: The Impact of Poverty, Food Insecurity, and Poor Nutrition on Health and Well-Being. Washington, D.C.: Food Research & Action Center; 2017. Available from: https://frac.org/research/resource-library/hunger-health-impact-poverty-food-insecurity-poor-nutrition-health-well.
<background-color:#5B96A2;bivertical-align:baseline;>FAO</background-color:#5B96A2;bivertical-align:baseline;>,<background-color:#5B96A2;bivertical-align:baseline;>IFAD and WFP</background-color:#5B96A2;bivertical-align:baseline;>.TheStateofFoodInsecurityintheWorld2013. Themultipledimensionsoffoodsecurity.Rome:FoodandAgricultureOrganizationoftheUnitedNations; 2013Available from: http://www.fao.org/3/i3434e/i3434e00.htm.
Roser M, Ritchie H. Hunger and Undernourishment. Oxford: University of Oxford; 2013. Available from: https://ourworldindata.org/hunger-and-undernourishment.
World Count. How Many People Die From Hunger Each Year. Denmark: TheWorldCounts; 2020. Available from: https://www.theworldcounts.com/challenges/people-and-poverty/hunger-and-obesity/how-many-people-die-from-hunger-each-year/story.
Gustafson S. Food Security Portal. Food Crises Increase Around the World: 2020 Global Report on Food Crises Released. 2020 Washington, DC: Food Security Portal. Available from: https://www.foodsecurityportal.org/blog/food-crises-increase-around-world-2020-global-report-food-crises-released.
SOS. Hunger and Food Scarceness in Africa. Washington, DC: SOS Children’s Villages; 2018. Available from: https://www.sos-usa.org/about-us/where-we-work/africa/hunger-in-africa.
Weiser SD, Palar K, Hatcher AM, Young SL, Frongillo EA. Food insecurity and health: a conceptual framework. In Food insecurity and public health. CRC Press. 2015; 23 – 50.
Stuff JE, Casey PH, Szeto KL, Gossett JM, Robbins JM, Simpson PM, et al. Household food insecurity is associated with adult health status. J Nutr. 2004;134(9):2330–5.
Dinour LM, Bergen D, Yeh MC. The food insecurity–obesity paradox: a review of the literature and the role food stamps may play. J Am Diet Assoc. 2007;107(11):1952–61.
Seligman HK, Davis TC, Schillinger D, Wolf MS. Food insecurity is associated with hypoglycemia and poor diabetes self-management in a low-income sample with diabetes. J Health Care Poor U. 2010;21(4):1227.
Seligman HK, Bindman AB, Vittinghoff E, Kanaya AM, Kushel MB. Food insecurity is associated with diabetes mellitus: results from the National Health Examination and Nutrition Examination Survey (NHANES) 1999–2002. J Gen Intern Med. 2007;22(7):1018–23.
Laraia B, Siega-Riz AM, Gundersen C. Household food insecurity is associated with self-reported pregravid weight status, gestational weight gain and pregnancy complications. J Am Diet Assoc. 2010;110:692–701.
Laraia BA, Siega-Riz AM, Gundersen C, Dole N. Psychosocial factors and socioeconomic indicators are associated with household food insecurity among pregnant women. J Nutr. 2006;136:177–82.
Whitaker RC, Phillips SM, Orzol SM. Food insecurity and the risks of depression and anxiety in mothers and behavior problems in their preschool-aged children. Pediatrics. 2006;118(3):e859–68.
Miller CL, Bangsberg DR, Tuller DM, Senkungu J, Kawuma A, Frongillo EA, et al. Food insecurity and sexual risk in an HIV endemic community in Uganda. AIDS Behav. 2011;15(7):1512–9.
Tsai AC, Bangsberg DR, Frongillo EA, Hunt PW, Muzoora C, Martin JN, et al. Food insecurity, depression and the modifying role of social support among people living with HIV/AIDS in rural Uganda. Soc Sci Med. 2012;74(12):2012–9.
Vogenthaler NS, Kushel MB, Hadley C, Frongillo EA, Riley ED, Bangsberg DR, et al. Food insecurity and risky sexual behaviors among homeless and marginally housed HIV-infected individuals in San Francisco. AIDS Behav. 2013;17(5):1688–93.
Baig-Ansari N, Rahbar MH, Bhutta ZA, Badruddin SH. Child's gender and household food insecurity are associated with stunting among young Pakistani children residing in urban squatter settlements. Food Nutr Bull. 2006;27(2):114–27.
Gundersen C, Kreider B. Bounding the effects of food insecurity on children’s health outcomes. J Health Econ. 2009;28(5):971–83.
De Hoyos RE, Sarafidis V. Testing for cross-sectional dependence in panel-data models. STATA J. 2006;6(4):482–96.
Pesaran MH. Estimation and inference in large heterogeneous panels with a multifactor error structure. Econometrica. 2006;74(4):967–1012.
Pesaran MH. A simple panel unit root test in the presence of cross-section dependence. J Appl Economet. 2007;22(2):265–312.
Breusch TS, Pagan AR. The Lagrange multiplier test and its applications to model specification in econometrics. Rev Econ Stud. 1980;47(1):239–53.
Pesaran MH. General diagnostic tests for cross-section dependence in panels. 2004; IZA Discussion Paper No. 1240, University of Cambridge, UK.
Baltagi BH, Feng Q, Kao C. A Lagrange Multiplier test for cross-sectional dependence in a fixed effects panel data model. J Econometrics. 2012;170(1):164–77.
Tugcu CT, Tiwari AK. Does renewable and/or non-renewable energy consumption matter for total factor productivity (TFP) growth? Evidence from the BRICS. Renew Sust Energ Rev. 2016;65:610–6.
Friedman M. The use of ranks to avoid the assumption of normality implicit in the analysis of variance. J Am Stat Assoc. 1937;32(200):675–701.
Frees EW. Assessing cross-sectional correlation in panel data. J Econometrics. 1995;69:393–414.
Frees EW. Longitudinal and panel data: analysis and applications in the social sciences. Cambridge University Press; 2004.
Im KS, Pesaran MH, Shin Y. Testing for Unit Roots in Heterogeneous Panels. J Econometrics. 2003;115(1):53–74.
Maddala GS, Wu S. A comparative study of unit root tests with panel data and a new simple test. Oxf B Econ Stat. 1999;61(S1):631–52.
Choi I. Unit Root Tests for Panel Data. J Int Money Financ. 2001;20:249–72.
Levin A, Lin CF, Chu CSJ. Unit root tests in panel data: asymptotic and finite-sample properties. J Econometrics. 2002;108(1):1–24.
Breitung J. The local power of some unit root tests for panel data, in: B. Baltagi, editor, Nonstationary Panels, Panel Cointegration, and Dynamic Panels. Adv Econom. 2000; 15(JAI): 161–178.
Hadri K. Testing for stationarity in heterogeneous panel data. Economet J. 2000;3(2):148–61.
Bai J, Ng S. A Panic Attack on Unit Roots and Cointegration. Econometrica. 2004;72:1127–77.
Chang Y. Non-linear IV Unit Root Tests in Panels with Cross-Sectional Dependency. J Econometrics. 2002;110:261–92.
Chang Y. Bootstrap Unit Root Tests in Panels with Cross-Sectional Dependency. J Econometrics. 2004;120:263–93.
Choi I. Combination unit root tests for cross-sectionally correlated panels. 2002.
Phillips PCB, Sul D. Dynamic Panel Estimation and Homogeneity Testing under Cross Section Dependence. Economet J. 2003;6:217–59.
Harris R, Sollis R. Applied time series modelling and forecasting. Wiley; 2003.
Smith LV, Leybourne S, Kim TH, Newbold P. More powerful panel data unit root tests with an application to mean reversion in real exchange rates. J Appl Economet. 2004;19(2):147–70.
Moon HR, Perron B. Testing for Unit Root in Panels with Dynamic Factors. J Econometrics. 2004;122:81–126.
Cerrato M, Sarantis N. A bootstrap panel unit root test under cross-sectional dependence, with an application to PPP. Comput Stat Data An. 2007;51(8):4028–37.
Palm FC, Smeekes S, Urbain JP. Cross-sectional dependence robust block bootstrap panel unit root tests. J Econometrics. 2011;163(1):85–104.
O’Connell PGJ. The Overvaluation of Purchasing Power Parity. J Int Econ. 1998;44:1–19.
Hurlin C, Mignon V. Une Synthèse des Tests de Racine Unitaire sur Données de Panel. Economie et Prévision. 2005; 3–4(169): 253–294.
Baltagi BH. Econometric Analysis of Panel Data, 3rd edition. 2008. Chichester: John Wiley & Sons.
Chudik A, Pesaran MH. Large Panel Data Models with Cross-Sectional Dependence: A Survey. 2015. In The Oxford Handbook of Panel Data, edited by B. H. Baltagi, 3–45. New York: Oxford University Press.
Eberhardt M, Presbitero AF. Public debt and growth: Heterogeneity and non-linearity. J Int Econ. 2015;97(1):45–58.
Hashiguchi Y, Hamori S. Small sample properties of CIPS panel unit root test under conditional and unconditional heteroscedasticity. MPRA Paper. 2010. No. 24053.
Choi I. Unit root tests for cross-sectionally correlated panels. In Econometric theory and practice: frontiers of analysis and applied research. 2006.
Albulescu CT, Pépin D, Tiwari AK. A re-examination of real interest parity in CEECs using ‘old’ and ‘new’ second‐generation panel unit root tests. B Econ Res. 2016;68(2):133–50.
Westerlund J. Testing for error correction in panel data. Oxf B Econ Stat. 2007;69(6):709–48.
Westerlund J. Edgerton DL. A Panel Bootstrap Cointegration Test. Econ Lett. 2007;97(3):185–90.
Westerlund J, Edgerton DL. A simple test for cointegration in dependent panels with structural breaks. Oxf B Econ Stat. 2008;70(5):665–704.
Groen JJJ, Kleibergen F. Likelihood-based cointegration analysis in panels of vector error-correction models. J Bus Econ Stat. 2003;21(2):295–318.
Westerlund J. Panel cointegration tests of the Fisher effect. J Appl Economet. 2008;23(2):193–233.
Gengenbach C, Urbain JP, Westerlund J. Error correction testing in panels with common stochastic trends. J Appl Economet. 2016;31(6):982–1004.
Banerjee A, Carrion-i‐Silvestre JL. Testing for panel cointegration using common correlated effects estimators. J Time Ser Anal. 2017;38(4):610–36.
McCoskey S, Kao C. A Residual-Based test of the Null of Cointegration in Panel Data. Economet Rev. 1998;17(1):57–84.
Banerjee A, Dolado J, Mestre R. Error-correction mechanism tests for cointegration in a single‐equation framework. J Time Ser Anal. 1998;19(3):267–83.
Abdullah SM, Siddiqua S, Huque R. Is health care a necessary or luxury product for Asian countries? An answer using panel approach. Health Econ Rev. 2017;7(1):1–12.
Martins PM. Aid absorption and spending in Africa: A panel cointegration approach. J Dev Stud. 2011;47(12):1925–53.
Pedroni P. Critical Values for Cointegration Tests in Heterogeneous Panels with Multiple Regressors. Oxf B Econ Stat. 1999; 61:653–70.
Pedroni P. Panel cointegration: asymptotic and finite sample properties of pooled time series tests with an application to the PPP hypothesis. Economet Theor. 2004;20(3):597–625.
Kao C. Spurious Regression and Residual-Based Tests for Cointegration in Panel Data. J Econometrics. 1999;90:1–44.
Driscoll JC, Kraay AC. Consistent covariance matrix estimation with spatially dependent panel data. Rev Econ Stat. 1998;80(4):549–60.
Hoechle D. Robust standard errors for panel regressions with cross-sectional dependence. STATA J. 2007;7(3):281–312.
Breitung J, Pesaran MH. Unit roots and cointegration in panels. In: The econometrics of panel data. Berlin: Springer; 2008.
Firebaugh G, Warner C, Massoglia M. Fixed, Effects. Random Effects, and Hybrid Models for Causal Analysis. In: Morgan S, editor. Handbook of Causal Analysis for Social Research. Dordrecht: Springer; 2013. pp. 113–32.
Dumitrescu EI, Hurlin C. Testing for Granger non-causality in heterogeneous panels. Econ Model. 2012;29(4):1450–60.
Bayar Y, Odabas H, Sasmaz MU, Ozturk OF. Corruption and shadow economy in transition economies of European Union countries: a panel cointegration and causality analysis. Econ Res-Ekon Istraz. 2018;31(1):1940–52.
Barbieri L. Panel cointegration tests: a survey. Panel Cointegration Tests: A Survey. 2008; 3–36.
Kao C, Chiang MH. On the estimation and inference of a cointegrated regression in panel data. In: Nonstationary panels, panel cointegration, and dynamic panels. Emerald Group Publishing Limited; 2001.

Download PDF

Journal Publication

published 15 Feb, 2023

Read the published version in BMC Public Health →

Reviews received at journal
18 Apr, 2022
Reviewers invited by journal
24 Mar, 2022
Editor invited by journal
11 Mar, 2022
Editor assigned by journal
10 Mar, 2022
First submitted to journal
10 Mar, 2022

You are reading this latest preprint version

The impact of food insecurity on health outcomes: Empirical evidence from sub-Saharan African countries

Status:

Journal Publication

Version 1

Abstract

Figures

Background

Theoretical And Conceptual Framework

Methods

Data type, sources and model specification

Basic panel econometric tests and their justifications

Cross-sectional dependence (CD)

Unit root test

Cointegration test

Estimation techniques, procedures and justifications

Results

Basic panel econometric tests

Long-run estimation results

Robustness checks

Discussion

Cross-sectional dependence test

Unit root tests

Cointegration tests

Long-run estimation results

Conclusions

Abbreviations

Declarations

References

Status:

Journal Publication

Version 1