1. Introduction
For instance, COVID-19 reversed approximately three years of global progress on extreme poverty reduction (
World Bank 2020;
UN 2022,
2023). In 2020, the number of people living in extreme poverty globally, defined as persons living on less than USD 2.15 per day in 2017, purchasing power parity, rose to 724 million, surpassing the pre-pandemic projection by about 90 million (
World Bank 2020;
UN 2022,
2023). Thus, COVID-19 pushed 93 million additional people into extreme poverty in 2020 (
World Bank 2020;
UN 2022;
2023). In 2022, about 9.2 percent, or 735 million of the world’s population, faced chronic hunger, which is 122 million more than in 2019 (
UN 2023). About 29.6 percent, or 2.4 billion of the world’s population, were moderately or severely food insecure. This estimate of food insecurity reflects an alarming 391 million more people than in 2019 (
UN 2023).
The United Nations and its development agencies note that the effects of COVID-19 are far from over. The organization warns that food security may acutely deteriorate further in at least twenty (20) countries, most of which are developing African and Asian countries (
FAO and WFP 2022;
UN 2022,
2023). The United Nations further warns that the 2030 Agenda for Sustainable Development is in grave jeopardy due to multiple cascading and intersecting crises predominated by COVID-19 (
UN 2022,
2023). Despite these stark projections of livelihood changes caused by the COVID-19 pandemic, varying disease burdens and the differential effectiveness of country’s mitigation policies to SARS-CoV-2 across the globe may have differential impact of COVID-19 on national, regional, and local food systems (
Dasgupta and Robinson 2022;
FAO and WFP 2022;
Giwa-Daramola and James 2023;
UN 2023).
Thus, even though COVID-19 has consistently resulted in job losses, income shortfalls, and food shortages, the impact of COVID-19 may vary with the different economic structure of countries and the effectiveness of the country’s mitigation policies (
Weder di Mauro and Baldwin 2020;
Haug et al. 2020;
Rose et al. 2021;
Trump and Linkov 2020). For instance, large, agrarian-based economies may face different impacts compared to largely industrialized economies (
Dasgupta and Robinson 2022;
FAO and WFP 2022). Also, some countries adopted stay-in-place restrictions or lockdowns, while others adopted alternative approaches (
Weder di Mauro and Baldwin 2020;
Haug et al. 2020;
Rose et al. 2021). Even among those countries that adopted lockdown COVID-19 anticontagion measures, there were variants, such as partial and complete lockdowns. Two of the unanswered socioeconomic questions that we address in this paper are as follows: what are the structural effects of COVID-19 on the livelihoods of food system actors, and what is the effect of the anticontagion measures on the livelihoods of food system actors?
We answer the above questions using data from farming system actors from Ghana. The Ghana COVID-19 management system provides a case insight into the effects of the COVID-19 anticontagion policy and the structural effects of COVID-19 on the resilience of the food system in a developing country context. Like most sub-Saharan African countries (such as Nigeria, Kenya, Ethiopia, South Africa, and others), at the early phase of the pandemic, selected urban areas in Ghana were under three weeks of mandatory lockdown, with strict policing (
Weder di Mauro and Baldwin 2020;
Giwa-Daramola and James 2023;
Haider et al. 2020;
Haug et al. 2020;
Rose et al. 2021;
Trump and Linkov 2020). In rural and peri-urban areas, the residents were expected to observe social distancing and other COVID-19 preventive protocols, but there were no COVID-19 lockdowns (
Haider et al. 2020).
This disparity in the spatial offerings of COVID-19 anticontagion policies across the country may have had varying effects on the livelihoods of the food system actors in the various spatial units. For instance, there were reports of rice piles stuck in northern Ghana, with less defined distribution channels, market, and storage. These disparities in the spatial offerings of the COVID-19 anticontagion policies were also adopted across several sub-Saharan African countries and other developed and developing regions (
Weder di Mauro and Baldwin 2020;
Giwa-Daramola and James 2023;
Haug et al. 2020;
Rose et al. 2021;
Trump and Linkov 2020). To date, we know very little about the cascading effects of the spatial COVID-19 policy disparities on the livelihoods of the food system actors in these spatial units. However, the food system innovations and consumer lifestyle changes caused by the COVID-19 policies and restrictions may remain, and may even have extenuating effects on consumer and producer behavior at the national and sub-national levels. It is, therefore, imperative to assess the cascading effect of the COVID-19 anticontagion policies on the livelihood of food system actors. In line with this discourse, we seek to understand the structural and location-specific effects of the COVID-19 policies on the incomes and food security of food system actors.
As of October 2020, global governments had invested about USD 12 trillion into COVID-19 economic recovery programs (
Markandya et al. 2021). However, the COVID-19 economic recovery programs did not sufficiently address food security and sustainability (
Markandya et al. 2021). New policies and programs are necessary in order to address the livelihood challenges and the inequalities and injustices across food systems that have emerged and will prevail post-COVID-19 (
Barrett 2020;
Gruère and Brooks 2021;
Laborde et al. 2020;
Markandya et al. 2021). Such food system programs are relevant, particularly for Ghana and sub-Saharan African countries, where agriculture remains a major engine for growth and an avenue for poverty reduction and food security improvement. Disruptions to the food system were caused by COVID-19, therefore, they may have implications for the nearly 60 percent of the population that are directly and indirectly employed in the sector and the overall economies of these countries. The continued disruption of the food system and supply chains and the weak policies and programs that poorly address food policy challenges may also have some implications for social unrest in the already-fragile sub-Saharan African region (
Barrett 2020;
Giwa-Daramola and James 2023). This research provides guidance and indications for the government of Ghana and other countries across the sub-region to effectively address the unwanted effects of COVID-19 on the food system. The results may also guide the ongoing and future anticontagion policy discourse, design, and implementation, particularly in Ghana. It may also help other, similar low- and middle-income countries to restore and maintain a sustainable food system at national and sub-national levels.
4. Discussion and Policy Considerations
While the worst of the COVID-19 pandemic appears to be over, the world is still reeling from its impacts. We analyzed the cascading effects of the COVID-19 anticontagion policies on the livelihoods of food system actors across different policy-induced spatial units. This analysis uses cross-sectional pooled-panel data from farming system actors in Ghana. Given the multilevel nature of the data in a developing country setting, we use machine learning techniques, such as the general additive model (GAM), subset regression, and generalized linear models. We consider a broad array of COVID-19-induced food system issues in Ghana, which also had serious policy implications for the rest of the sub-Saharan African region, and utilize robust machine learning modeling approaches in this study to link food systems and COVID-19 policies.
We found that in the pandemic era, the average monthly income of the food system actors significantly decreased by about 11.6 percent. The decrease in income resulted from the closure of food and food-related businesses during the pandemic. We also tested the livelihood changes across three spatial units—the locked-down areas, the neighboring communities to the locked-down areas, and the communities that were extremely distant from the locked-down areas. We found that the COVID-19 mitigation policies led to cascading spatial economic variations among the food system actors across the spatial units. The income levels of the farming system actors in the lockdown areas, and those in the non-contiguous communities to the lockdown areas decreased, whereas the incomes of the farming system actors in communities directly adjacent to the lockdown areas increased. Overall, we found that the main losers were the farming system actors in the locked-down areas and those living in the distant communities that were non-contiguous to the locked-down areas. The main beneficiaries of the lockdown policies were the food system actors in the neighboring communities of the locked-down areas.
Our explanation for these differential incomes is that this income shifts stems from the travel restrictions, roadblocks, increased military, and other security agency presence, the fear of military brutality, and the closure of businesses. Due to the increased roadblocks and the fear of brutality due to the increased military presence in the locked-down areas, many of the food system actors, such as food transport truck operators, in the locked-down areas failed to operate. Also, because most of the locked-down communities tended to be mostly urban, most of the food system actors in these areas were traders, agro-processers, restaurant and small food joint operators, and agro-input dealers. With markets, restaurants, and businesses closed, and many of their clients at home due to lockdown, the demand for the services of the urban food system actors became diminished, leading to a reduced income. In a similar manner, even though the food system actors operating in the communities in distant areas that were not under lockdown also lost incomes, it was not as much as that of the actors in the locked-down areas. Most of these distant communities tended to be rural and, as such, relied on derived urban food demand to gain income. In fear of harassment at roadblocks in the locked-down areas, the actors in this spatial unit could not supply their farm produce or their services to the urban locked-down areas for income. On the contrary, the farming system actors in the contiguous peri-urban or suburban areas to the locked-down areas took advantage of the artificial food supply shortages induced by limited food supply from the rural distant areas. Those farming system actors gained more income by supplying food and food system services to clients both in their community and in the locked-down areas. The increased demand for their goods and services, due to the supply gaps from the rural distant communities, also led to price hikes in the food commodities and services, which contributed to an increased income for the food system actors in the communities bordering the locked-down areas.
The levels of food security across the three spatial units of the COVID-19 policies also followed similar patterns as that of the income distribution. The level of food security was reduced in both the lockdown areas and the distant rural communities, but less so in the adjoining rural communities. In the adjoining communities, the levels of food security increased. The food supply shortages, the disruption of traditional markets (which involved in some cases a complete lockdown of markets for fumigation), or rationing in market use explain the rise in food insecurity in the locked-down areas. In the distant non-contiguous communities, the decrease in food security may be the result of shortages in industrial food products and associated price hikes.
Even though there were general increases in food expenditure and food expenditure share across all the spatial units of the farming system actors, we see relatively high food expenditure among the actors in both the direct locked-down contiguous areas and the non-contiguous areas. The increased food expenditure in these two spatial units might have resulted from a reduced supply of processed food items from the urban locked-down areas to the adjoining suburban and rural areas, due to disrupted transportation from increased security posts and roadblocks. The increased food expenditure, especially in the non-contiguous areas, combined with the reduced income, reflects in the highest food budget share of the farming system actors in the non-contiguous locked-down areas. In the locked-down and adjoining areas, the food system actors increased their food expenditure shares, relative to the distant non-locked-down non-adjoining areas. The increasing food demand and high prices in most instances led to high food insecurity, especially among the existing vulnerable households (
Huss et al. 2021;
Apaliya et al. 2022;
Asravor and Kwakwa 2022).
We also found major changes in the behavioral patterns and a deterioration in the livelihood outcomes across the general food system structures. On average, the respondents expended an additional 14 percent of their income on food purchases. This increase in food expenditure constituted an increase of about 42.7 percent in the share of the respondent’s income expended on food purchases, relative to the pre-pandemic period. The adult food security estimates decreased by 13.8 percent, from 80 percent in the year before the pandemic to 69 percent during and after the pandemic. The child food insecurity, likewise, decreased by about 10.3 percent, from the pre-pandemic level of 87 percent to a pandemic level of about 78 percent. Adult food rationing may have contributed to the relatively lower decreases in child food security and the higher increases in adult food insecurity. Full-time employment decreased by 14.6 percent, from 84.9 percent in 2019 to 72.5 percent in 2020, whereas part-time employment increased by about 82.1 percent, from 15.1 percent in 2019 to 27.5 percent in 2020. Due to the increases in part-time jobs, we found that the average number of months that the unemployed respondents were out of employment decreased from 10.6 months in 2019 to 9.1 months in 2020, a decrease of about 14.8 percent in unemployment months.
Even though our findings have many important policy implications for the design and implementation of food system policies and programs for low- and middle-income countries, we settle on two key policy issues. First, effective post-pandemic food system policymaking and programs demand matching policy incentives for food systems support and services with the cascading and trickling down effects of the COVID-19 anticontagion policies. From the perspective of the cost-effective design of food policies and programs, understanding the spatial dimensions of the COVID-19 policies that influence the food system actor’s livelihood outcomes and behavioral patterns can inform policymakers about heterogeneity in the effect of the pandemic policies on food system actors. Such knowledge contributes to the creation of tailored policies towards food system actors’ heterogeneous preferences and the spatial targeting of policy interventions. In this study, we have found that the lockdown policies adopted during COVID-19 have created heterogenous food system effects on livelihood outcomes that blanket policies may not address.
We also note that the Ghanaian government has expended over USD 100 million on COVID-19 recovery programs, and, globally, over USD 12 trillion has already been spent on COVID-19 economic recovery programs (
Markandya et al. 2021;
Ministry of Finance 2022;
World Bank 2023). While we do not know the specific impacts of the COVID-19 recovery programs on the Ghanaian food system, the global estimates from the United Nations show that more than 4 billion individuals of the world’s vulnerable population groups, including the young and the elderly, remain uncovered by the statutory social protection programs (
UN 2023). Also, the COVID-19 economic recovery programs have not sufficiently addressed food security and sustainability (
Markandya et al. 2021). The United Nations warns that the economic recovery from the COVID-19 pandemic has been slow and uneven, with 41 percent of low-income countries experiencing a higher poverty rate in 2021 compared to the previous year, versus only 13 per cent of the upper-middle-income countries (
UN 2023). The United Nations stress that the Sustainable Development Goals are in grave jeopardy; moreover, under the current recovery trends, 575 million people will still be living in extreme poverty in 2030, and only about one third of countries will meet the target to halve the national poverty levels. The world is back at hunger levels that have not been seen since 2005, food prices remain higher in more countries than in the period of 2015–2019, and the lingering effects of COVID-19 continue to derail the progress in food security. Our findings do not only give insight into the most vulnerable population, but also the locations of the most vulnerable groups for effective policy planning and targeting. New policies and programs are necessary to address the livelihood challenges and inequalities and injustices across the food systems that have emerged and prevail post-COVID-19 (
Barrett 2020;
Gruère and Brooks 2021;
Laborde et al. 2020;
Markandya et al. 2021)
Second, even though the COVID-19 anticontagion policies have been largely successful for mitigating the pandemic, there are lingering concerns regarding the effects of the policies on food system actors, which can be used to guide future pandemic policies. We address this concern in our paper. The effects of COVID-19 on food system actors vary, not only based on their resilience, but also on the nature of the spatial policies adopted to curb the spread of the disease. Our paper lends support to effective anticontagion policies that assure the free movement of goods and services, devoid of military interventions and roadblocks, and combine continuous effective communication.
5. Conclusions and Recommendations
Even though sustained episodes of growth are underpinned by fundamental structural changes, the full effects of COVID-19 are devasting and may have extenuating effects over a long period of time. Understanding these effects and the role of the lockdowns is important for shaping future pandemic management. From this analysis, we can see that COVID-19 shifted the structure of local economies, and food systems are among those affected. The food system actors in the locked-down areas and the distant areas from the locked-down communities are the most affected groups. This suggests that the lockdown policies may have contributed to the observed low income, employment shifts, and food insecurities among the farming system actors in these areas. The government of Ghana, through funding support from the World Bank, established the COVID-19 alleviation and revitalization of enterprises support (Ghana CARES Obaatanpa) program to mitigate the impact of COVID-19 on livelihoods and to ensure that Ghanaians can quickly emerge from the pandemic (
Ministry of Finance 2022). The government may still need to do more to balance out the disproportionate effects of the anticontagion polices across the different spatial units. For instance, food inflation—which was brought stark by COVID-19, during which local food prices increased in some parts of the country—has remained, and even elevated to about 40.1% in August of 2023. The government, therefore, needs to provide additionaleconomic support and assitance to the food system actors across the policy areas. Government support to food system actors will have micro-level benefits of enhancing the livelihoods of the food system actors, and macro-level implications of reducing food inflation.
Our study is subject to several limitations. First, our empirical analysis is only limited to one country. Even though the results from this analysis can be indicative of the outcomes in similar countries, it is not entirely representative of the entirety of sub-Saharan Africa. Second, the period of study is short, and thus, the outcome may not fully capture the dynamic changes in the livelihoods of food system actors. Considering that COVID-19 effects still linger on Ghana’s economy and the food system, as evidenced through high food inflation and the full-blown macro-economic crisis in 2022, studies to track the trajectory of the COVID-19 impact over a long period will be useful (
World Bank 2023;
OBG 2023). We are unable to identify a meaningful instrument for COVID-19. While we tested for correlation, multicollinearity, and endogeneity and used robust standard deviations in our models, future studies that can identify the instruments for COVID-19 may give new insights. However, it is important to consider that socioeconomic events, including COVID-19 policy effects, are complex, far-reaching, and multidimensional. Therefore, it is harder to come up with complete, well-defined models in such a non-experimental natural environment, making it more difficult to predict than in the physical process or framed field experimental settings. It is important to consider other metrics of model quality, such as its predictive capacity in test data and error margins in a machine learning environment, in order to gain a more comprehensive understanding of the performance of the models (
Giwa-Daramola and James 2023). Our results have extremely low test errors of approximately 11 percent, which means that the model accurately predicts outcomes around 90 percent of the time. Notwithstanding the limitations of this study, we have drawn important conclusions about the micro-level effects of the pandemic policies on the livelihoods of the food system actors in a sub-Saharan African context. Like Ghana, most of these countries are still facing the lingering effects of COVID-19 across their food systems, which governments and other policymakers must urgently address.