The impact of the COVID-19 pandemic on vegetable farmers in Bangladesh

Abstract

The COVID-19 pandemic not only imposed severe health risks but also raised major challenges to the economy, due to widespread and severe measures to control the spread of the disease. Food value chains were disrupted by restrictions of the movement of people and commodities, which had significant impacts on the livelihoods of smallholder farmers. The purpose of this research is to determine the impact of the COVID-19 pandemic on Bangladeshi vegetable farmers. A total of 320 vegetable farmers were selected from the North-West region of Bangladesh. Both quantitative and qualitative data were collected through a digital survey method. Analysis revealed that around 3–5% of the marketed surplus of the farms was reduced during the pandemic due to the disturbances. The majority of the farm households reported that there was a significant reduction in their family income and, as a consequence, around 38% of farm households had cut down on their food consumption during the pandemic. The farmers were found to follow different strategies to cope with the difficulties and respond to government initiatives to mitigate such impacts. Despite all the restrictions and risks, extension services were still available to help the farmers. On the basis of the findings, this study suggests the importance of collaborative participation of the relevant bodies to decrease the effects of COVID-19 on farm households by employing all available mechanisms and focusing more on identifying effective coping strategies that can be supported in the event of future shocks, for more sustainable and resilient food systems.

Keywords:

• COVID-19
• vegetable farmers
• livelihoods
• impact
• north-west region
• Bangladesh

1. Introduction

In late 2019 the COVID-19 outbreak began, which would eventually reach pandemic proportions and cause not only human health risks but also major disruptions to local and global food systems. Governments around the world responded to the threat by restricting the movement of people, which resulted in disruptions to the movement of commodities. In Bangladesh, COVID-19 was first identified in March 2020. Initially, the pandemic’s progress was slow, but in mid-June 2020, the number of infected patients was found to be over 100,000 (Nath et al., 2020). Since then, estimated 2 million people have been infected in Bangladesh and 29,200 people have died (WHO, 2020). Bangladesh was compelled to face lock-downs at various times and stages as the number of COVID-19 cases rose every day. The first day of the closure and lockdown was 26 March 2020 (Shawon and Mamun 2020) and a transport ban was announced on all modes of transport across the country from 26 March 2020 to 11 April 2020. Then Bangladesh proclaimed a second state of emergency on 14 April 2021. Overall, Bangladesh was in a national lockdown from March to May 2020, which drastically disrupted food value chains by restricting the movement of people and commodities, including agricultural inputs and outputs, resulting in food loss and waste across the country (Barrett, 2020; Islam et al., 2021; Roy, 2020). In addition to damaging public health, COVID-19 had a significant impact on a number of economic sectors, including agriculture, food security, food supply, trading activities, transportation system, and industry in Bangladesh (Begum et al., 2020).

The Bangladeshi agricultural industry is the key source of food supply throughout the country, and the pandemic is anticipated to cost the government roughly USD 625 million (Bangladesh National Nutrition Council, 2020; Mishra & Rampal, 2020; Termeer et al., 2020). During the pandemic, every sector of agriculture (crops, livestock, and fisheries) in Bangladesh experienced serious problems. The pandemic occurred at a time when farmers were preparing for the Rabi crop harvest, and the lockdown continued during the production time of Kharif-1 crops. As a result, farmers faced challenges in obtaining labor, extension services, as well as providing supplies (seeds, fertilizers, insecticides, and labour) for new cropping seasons. Agriculture’s most vulnerable sector is perishable items, such as fresh fruits and vegetables. The unexpected shutdown caused supply chain disruptions, notably in the case of vegetables, which are high-value perishable crops. Local marketplaces have timing limits and within a certain time, farmers cannot sell their produce. Most vegetables deteriorate quickly, so farmers must sell them at a loss. Cucumbers, eggplants, and tomatoes were sold for 300–400% less than average (Ubinig, 2020). On the other hand, production costs were higher due to challenges with agricultural inputs, labor shortages, and transportation. Farmers experienced difficulty harvesting crops or bringing their products to market as a result of the new mobility limitations (Siche, 2020; Zhang et al., 2020). The number of automobiles on the road significantly decreased. As a result, the cost of food transportation rose (FAO, 2020a). Labor shortages hampered agricultural productivity, and travel limitations hindered both sellers and purchasers’ access to marketplaces (Zabir et al., 2020). Urban workers who lost their employment and migrant workers from abroad returned home to their rural villages creating a labor surplus. At the same time, would-be migrant workers were restricted to travel to village communities creating labor shortages (Kabir et al., 2021; Karim et al., 2020; P. Kumar et al., 2021). Farmers were also facing serious financial challenges as a result of declining product demand, supply disruptions, and being forced to sell below cost (Termeer et al., 2020).

Many people in poor nations, such as Bangladesh, lost their jobs as a result of the pandemic, and their purchasing power decreased. As a result, demand for high-priced goods fell. Prices for agricultural items frequently fluctuated depending on commodity and locality. For example, prices for perishable goods such as vegetables, fish, and chicken dropped sharply due to a lack of buyers and traders in local markets or rumors about food safety concerns (Alam et al., 2021; Sunny et al., 2021; Zabir et al., 2021); in contrast, prices for main food commodities such as rice and fish rose sharply in urban areas (Khatun et al., 2020 Mandal et al., 2021). Food loss during the pandemic was additionally compounded by rising transportation costs and insufficient storage facilities. Due to a lack of proper storage facilities, large volumes of eggs, milk, fruits, and vegetables were thrown away (Termeer et al., 2020). The consumer, on the other hand, paid special attention to vitamin C-rich fruits, which can help to improve the immune system (Khan et al., 2021).

These consequences affected both local and large-scale agriculture and slowed food distribution systems. Unfortunately, the effect of such crises will continue to have a detrimental impact on the poor and vulnerable communities. According to the FAO, severe negative effects on producers, transporters, processors, and consumers have already been recorded and will continue (FAO, 2020b). Prior to COVID-19 underdeveloped and disadvantaged countries like Bangladesh were already dealing with unreliable food supply systems that disrupted even slight food security issues (Béné, 2020). Unfortunately, the pandemic threatened the Sustainable Development Goals, particularly the two food-security-related goals of no poverty and zero hunger, which were badly hit during the lockdown period, especially in developing nations (Workie et al., 2020). Because of COVID-19, national governments enacted a variety of movement restrictions in an attempt to mitigate the pandemic’s impact, disrupting domestic and global agricultural value chains (Barrett, 2020; Elleby et al., 2020). As a result, global policymakers projected COVID-19 to have a significant impact on productivity and food security, particularly among the poor in rural areas (UNESCAP, 2020). COVID-19’s impacts on reduced agricultural production and food and nutrition security for regions such as sub-Saharan Africa (Ayanlade & Radeny, 2020), the Caribbean (Blazy et al., 2021), the Pacific Islands (Iese et al., 2021), or individual countries such as Bangladesh (Mandal et al., 2021), China (Pu & Zhong, 2020), Peru (Vargas et al., 2021) and Nepal (Adhikari et al., 2021).

Bangladesh is suitable for the production of a variety of vegetables, with over 142 types of locally grown and exotic vegetables produced in the country (BARI, 2018) and vegetables are grown during both the winter (mid-December to mid-February) and summer (mid-April to mid-June) seasons. In Bangladesh approximately 16.2 million farm households produce vegetables commercially, primary smallholder farm households with 0.05–2.49 acres of land, accounting for around 2.63% of total cultivable land (BARI, 2018; DAE, 2018; Weinberger & Genova-II, 2005). Due to the COVID-19 pandemic, the vegetable supply chain was disrupted, resulting in significant losses for farmers. Despite the fact that agricultural stimulus packages were published, they did not provide clear incentives for smallholder vegetable producers who have lost a lot of money on their crop (Alam et al., 2021). This raises the question of vegetable growers’ vulnerability in terms of farming, livelihoods, and food insecurity. To mitigate the sudden diverse effects of such a pandemic, farmers tried to adjust their own strategies and their agricultural practices including selling their assets, reducing food consumption, diversifying their sources of income, limiting the household expenditures, planting fewer crops, reducing their dependence on inputs and cropped area cultivated, or increasing labour in agriculture (Mahmud & Riley, 2021; Nolte et al., 2022; Obayelu et al., 2021) to cope with the impacts. So, long-term welfare implications of the coping methods used by farming households during shocks should be understood. Policymakers and related stakeholders need evidence of the effects of COVID-19, their adaptation strategies, and the success of government intervention programs to aid them in establishing effective intervention programs for future shocks.

Our hypothesis based on previous research is that vegetable production might be disrupted due to the problems faced in input supply, and sales might fall for these farmers due to the problems faced in accessing the market, impacting their livelihoods. Thus, reduced incomes are likely to lead to different coping mechanisms that might have a positive or negative impact on the farm households. The main objective is to assess the impact of the COVID-19 pandemic on farmers engaged in high-value agriculture in North-West Bangladesh, by exploring how COVID-19 affected vegetable farmers in terms of their access to input and output markets, the impact on their sales, the coping strategies they followed to deal with the impact of the disruption, and the role of government initiatives in mitigating the impact of the pandemic for vegetable farmers. It is imperative to conduct an impact evaluation and establish frameworks for shock mitigation to comprehend the ability of farmers in response to various types of shocks. Generating knowledge on individuals’ perceptions and resulting behavioral responses is crucial in designing appropriate policy measures, particularly for social protection. Typically, individuals engage in a cognitive evaluation of the shock experience as they initiate coping and response mechanisms. Thus, this research will have the potential to offer valuable perspectives on the vulnerabilities of the agricultural sector in times of pandemics and other crises. The aforementioned disruptions can potentially aid in the formulation of policies and strategies aimed at alleviating their impact on the agricultural sector. Through an examination of the coping strategies implemented by farmers amidst the pandemic, this study will be useful to identify the efficacious and sustainable approaches for managing similar disruptions in the future. Finally, this research has the potential to make a valuable contribution to the existing body of literature on resilience and adaptive capacity when dealing with crises.

The remainder of the paper is organised as follows: Section 2 presents the review of literature; Section 3 presents the methodology followed by the results and discussions in section 4. The last section presents conclusions and recommendations.

2. Literature review

Several studies have tried to analyze the implications of the COVID-19 pandemic on agricultural production and supply chains globally (A. Kumar et al., 2020; G. D. Sharma et al., 2020; Galanakis, 2020; Gillespie, 2020; Gregorioa & Ancog, 2020; Pu & Zhong, 2020; Singh & Neog, 2020; Solomon et al., 2020; Weersink et al., 2020; Workie et al., 2020; Zhang et al., 2020). Gregorioa and Ancog (2020) stated that the COVID-19 pandemic resulted in a reduction of 17.03 million tons of agricultural production in Southeast Asia. A. Kumar et al. (2020) argued that the vulnerabilities in agricultural supply chains and depleted workforces caused by the COVID-19 crisis have affected all types of farms and emphasized on the need for agricultural market reforms, digital solutions and safety nets to make them more resilient. For instance, Bochtis et al. (2020) showed that 50% of the agricultural workforce and 54% of workers’ yearly income are at moderate-to-high risk due to the COVID-19 pandemic, which has had a negative impact on wage rate and unemployment conditions, demonstrating the economic uncertainty of rural people (Sinha, 2021).

According to Barrett (2020), supply chain disruptions in agri-food systems occurred as a result of movement restrictions during COVID-19, which impeded the efficient flow of agricultural inputs and outputs. This pandemic has affected horticulture and food crop farming activities by limiting access to inputs, reducing yields, reducing income, limiting access to local and urban markets, restricting access to transportation, and increasing post-harvest losses, all of which put food security at risk (Khan et al., 2021; Middendorf et al., 2022; Poudel et al., 2020; Pulubuhu et al., 2020; Wulandari et al., 2022). Again, supply chain constraints prevent agricultural products from reaching consumers, and to counteract this, farmers have lowered their prices by incurring losses (Debata et al., 2020). Ali and Khan (2020) investigated the impact of the COVID-19 lockdown on wholesale prices of fruit and vegetables during the lockdown in India and discovered that wholesale prices of more perishable fresh fruits and vegetables with a higher water content decreased. Harris et al. (2020) and Siddiquei and Khan (2021) also found that the shutdown of large wholesale vegetable marketplaces lowered wholesale market prices and reduced farmers’ purchasing power, influencing their consumption of diversified foods. The delay in market supply led farmers to either damage crops or sell them at cheaper prices, which likely did not cover their production costs. Similarly, in Bangladesh, Zabir et al. (2021) found crop harvesting was at risk because most farmers were not mechanized and had labor shortages. Perishable products were priced low due to buyer shortage and supply channel constraints. Since the COVID-19 prognosis seemed unknown, they also anticipated it would take years to recover from this shock. According to Alam et al. (2021), during COVID-19, the brinjal, cucumber, pointed gourd, yard long beans, and bottle gourd lost around Tk. 4,900, Tk. 10900, Tk. 57400, Tk. 52500, and Tk. 18500 per acre, respectively, resulting in a drop in the price of produce by more than half. Vegetable farmers had taken a major financial hit, which then raised the specter of their food insecurity.

Thus, it became a subject of concern how the COVID-19 pandemic would affect rural living in terms of access to healthcare, food and nutrition, and livelihood opportunities (Janssens et al., 2021; Modak et al., 2020; Wang et al., 2021). The effects of the COVID-19 pandemic on food security, income, and livelihood, as well as the assessment of social support programs implemented during the pandemic to reduce the risk to vulnerable populations, have been the subject of numerous academic research studies (Barman et al., 2021; Bhorat et al., 2021; Brown & Cowling, 2021; Gu & Wang, 2020; J. Sharma et al., 2021). Adebiyi et al. (2021) discovered a moderately detrimental effect of COVID-19 on farms as a result of coping mechanisms used at the time. Middendorf et al. (2022) showed that food insecurity and a lack of social supports were key difficulties for resilience and livelihoods in smallholder farm households, especially for women, and for young people, who experienced increased unemployment and worries about poverty. Differential effects on smaller and larger farms, as shown by Harris et al. (2020), suggest that different sizes of farms may require different sorts of support to continue operating, while many farms maintained sales with a variety of coping strategies. Ali and Khan (2020) looked at what factors influence whether or not rural households received aid through social assistance programs, and they made recommendations for how to better help those living in rural areas in times of crisis. It was demonstrated by Agboola et al. (2021) that the government prioritized large businesses over individual farmers while aiding during the COVID-19 pandemic.

Studies from around the world demonstrate that the COVID-19 pandemic has negatively impacted vegetable growers by reducing yields and profitability, disrupting supply chains, and limiting market access. Food insecurity and rising poverty among rural smallholders have also resulted from the pandemic. Assistance programs implemented by the government during the pandemic have also been questioned. However, few studies have been conducted in Bangladesh that capture the overall understanding of being impacted by COVID-19, particularly in terms of identifying the specific problems that farmers faced in their input and output market and the empirical analysis of those problems on their sales, their action to resolve those problems, along with their strategies and government supports to mitigate the impact on their livelihood. The findings of this study will help us better understand the consequences of this shock in the country’s agriculture sector and will help the think tanks to depict the policy measures to build resilience for the future.

3. Methodology

3.1. Data

This study is conducted in the North-West region of Bangladesh. A multi-stage sampling procedure was followed to collect necessary data. Firstly, two division—Rajshahi and Rangpur were selected which comprise the North-West region of Bangladesh. In the second stage four districts—Bogura, Pabna, Rangpur and Dinajpur were selected based on the suitability of vegetable production and lastly, based on the availability of vegetable farmers, a total of eight upazilas (an administrative unit that serves as a sub-district) were selected from those four districts. A list of farmers involved in vegetable production was collected from each upazila agriculture office. Then, for each upazila, 40 vegetable farmers were selected randomly from that list (Table 1). Thus, a total of 320 vegetable farmers were surveyed to achieve the objectives. The data was collected from April to May 2022.

Table 1. Study area and sample size

Divisions	Districts	Upazilas	Sample size
Rajshahi	Bogura	Shibganj	40
		Sherpur	40
	Pabna	Ishwardi	40
		Atghoria	40
Rangpur	Rangpur	Pirganj	40
		Mithapukur	40
	Dinajpur	Parbatipur	40
		Chirirbandar	40
Total			320

A face-to-face interview questionnaire was developed to collect the survey data. Before finalizing the questionnaire, it was pre-tested. The questionnaire was translated into the Bengali language in order to obtain appropriate comments from respondents. The final version of the questionnaire was imported in a web-based digital data collection tool, mWater portal. Both qualitative and quantitative data were collected for this study. Several questions were asked to identify comparative information on the situation before a pandemic and during a pandemic, and in those cases, we used the year 2019 as the before pandemic period and the year 2020 as the pandemic period. Thus, we get an unbalanced panel data of total 639 observations, as one of the respondents could not provide some information and dropped from the panel dataset.

The interview schedule sought information based on the modules of socio-demographic profile, household income, expenditure, cost, and returns from vegetable farming, problems faced in accessing input and output markets, strategies followed by the farmers and government initiatives to mitigate the impact of pandemic. The data analysis methods are both descriptive and econometric. Descriptive analyses are followed to identify the socio-demographic status of the respondents, to identify the problems faced by the farmers in accessing both input and output market, to identify the strategies followed by the farmers to mitigate the impact of the pandemic, and to identify the government incentives to mitigate the effects of the pandemic on vegetable’s farmers.

3.2. Analyzing the input and output market problems faced by the vegetable farmers

Problem Confrontation Index (PCI) is used for identifying the problems faced by the farmers to access into agricultural inputs (seed, fertilizer, labour, pesticides, irrigation etc.) and marketing of their outputs (vegetables produced for sale) during the pandemic. Farmers were asked to give their opinion on the severity of selected problem items which were faced during the COVID-19 pandemic time period, by highlighting the correct response among the four alternatives as either “severe problem”, “moderate problem”, “low problem”, or “no problem”. Weights were assigned to these alternative responses as 3, 2, 1, and 0, respectively. The PCI of input market was determined according to the following equation:

PCI = 3*fs +2*fm +1*fl +0*fn

Where fs= number of respondents facing a severe problem.

fm = number of respondents facing a moderate problem.

fl = number of respondents facing a low problem.

fn = number of respondents facing no problem.

According to the equation, the value of PCI will depend on the number of respondents (fs, fm, fl and fn) and their relevant responses among the four alternatives- “severe problem”, “moderate problem”, “low problem”, or “no problem” (3, 2, 1, or 0). So, the actual range of PCI depends on the sample size of a study. In our case, we may say that the value of PCI could be ranged from 0 (if all of the farmers reported that they were facing no problem) to 960 (if all the farmers reported that they had were facing severe problem).

3.3. Impact of COVID-19 on market access of vegetable farmers

An econometric analysis was conducted to find out the impact of COVID-19 on the market access of vegetable farmers. The outcome variable is marketed surplus which gives the actual quantity of the produce sold by the farmers, considered as a proxy variable of market access.

MT = MS + PS—L

Where MT is the marketed surplus, MS is the marketable surplus, PS is the past stock sold out, if any, and L are the losses during storage and transit of the marketable surplus left for sale.

This study used Propensity Score Matching (PSM) to analyze the data. PSM is a quasi-experimental technique where the researcher matches each treated unit with a non-treated unit with similar features to create an artificial control group using statistical tools. PSM, in this instance, uses observed characteristics to determine the likelihood that a unit will be involved in a program or event. PSM compares treated units against untreated ones. This method is predicated on the idea that, subject to certain observable characteristics, treated and untreated units can be contrasted, as if the treatment had been entirely randomized. In order to avoid the problems of selection bias that afflict non-experimental approaches, PSM attempts to simulate randomization. The benefit of using PSM is, it has the ability to convert a multivariate into an index, namely a propensity score (PS), and the PS value can be utilized to match the control and treatment groups. This significantly lowers self-selection and confounding bias, allowing for more accurate treatment outcomes (Sánchez-Toledano et al., 2018)

In this study, we used the PSM method for analyzing the impact of COVID-19 by comparing the same farmers before and during pandemic as the PSM method is well suited for the analysis of such observational studies. In this case, the farmers are considered as the treatment group when they were affected by COVID-19 (2020), while the same farmers are considered as the control group when they were not affected by COVID-19. More specifically, marketed surplus of vegetables during COVID-19 is considered as the treatment group, and the marketed surplus of vegetables before COVID-19 is considered as a control group. PSM matches the treatment group (marketed surplus of vegetables during COVID-19) and control group (marketed surplus of vegetables before COVID-19) based on observed characteristics and identified problems to estimate the impact of the pandemic on the market access ability of the vegetable farmers.

Two steps were followed to estimate the average treatment effect on the treated (ATT). First, a binary probit model was estimated to calculate the PS which is the conditional probability of each sample to be affected by the COVID-19 pandemic (Habiyaremye, 2017; Liu et al., 2021). The binary probit model is as follows:

$Yi=Xi\beta i+ui,$

With $Yi$ = 1 (year = 2020) or $Yi$ = 0 (year = 2019). Here, $Y_i=$ the dependent variable denoting percentage of marketed surplus of vegetables. ${\beta }_i$are the parameters to be estimated, $X_i$ are the explanatory variables and $ui$ is the error term.

In the second step, based on the estimated PS, treatment group and control group were matched, and the mean difference of outcome was considered the impact of COVID-19 pandemic on marketed surplus. This ensured that the primary characteristics of the control and experimental groups were as comparable as possible. The balance requirement was assessed to identify whether statistically significant differences between the two groups persisted after resampling. This would ensure that the matching procedures balanced the data and achieved the effect of a randomized experimental design (Liu et al., 2021).

For PSM, there are various matching methods to handle selection bias. The most common method is balancing covariates to handle selection bias in PSM. This study has followed four matching algorithms nearest neighbor, radius matching, kernel matching and stratification matching to calculate ATT values in the treatment and control groups. The ATT values and impact of the COVID-19 pandemic on the control and treated groups after matching were compared (Li et al., 2015; Liu et al., 2021).

$ATT=E\left(Y1i-Y0i\right)=E\left\{E\left(Y1i-Y0i\right)\right\}=E\left\{E\left(Y1i\right)-E\left(Y0i\right)\right\}$

Here, $Y1i$ and $Y0i$ represent the percentage of marketed surplus of vegetables in the treated group and the control group, respectively.

3.4. Analyzing the coping strategies and government supports

In the context of analyzing the coping strategies followed by the farm households and also the government support provided to mitigate the impact of COVID-19, we used descriptive measures to represent the scenario. The results were represented in terms of identifying the strategies followed by the farm households to cope with the COVID-19, changes in income and expenditure of the farm households, change in consumption of different food groups, change in sources of inputs of farmers, change in labour participation, use of different sources of credit to take loan, and govt. incentives received by the rural households during the pandemic period.

4. Results and discussion

4.1. Descriptive statistics

From the primary analysis, this study has found that the respondents consisted of 92% males and 8% females. In terms of age group, the majority were 31–40 (27%) years of age; those aged 21–30 accounted for 18%, those aged 41–50 25%, and those above 51 25% of the respondents. The main occupation of the respondents were agriculture and average farm size was 1.8 acres. Considering the farm size, 77% of farm households consisted of small farm (<2.5 acres), 22% of medium farms (2.5 acres −7.49 acres) and only 1% were large farms (>7.49 acres) among the 320 farm households.

Table 2 shows that the average age of the respondents was 43.26 years. The average number of family members in farm households was five. Around 80% of farmers were found literate at different levels of education (primary, secondary, higher secondary, and upper). The market distance was found to be an average of 1.95 km for the households. The monthly income and expenditure of farm households decreased significantly during the pandemic year (2020) compared to the previous year (2019), because the majority of farmers (58–81%) reported that they were significantly affected by various problems related to input and output markets during the pandemic.

Table 2. Descriptive statistics of some important socio-demographic variables

Variables	Before Covid (2019)		During Covid (2020)		Mean Difference (2020–2019)
	Mean	SD	Mean	SD
Age (Years)	43.279	0.747	44.263	0.745	0.984
Primary (Dummy)	0.313	0.465	0.319	0.467	0.005
Secondary (Dummy)	0.288	0.454	0.287	0.453	−0.001
Higher secondary and upper (Dummy)	0.207	0.406	0.206	0.405	−0.001
Total monthly income (Tk)	25,430.88	1,129.68	18,953.16	839.28	−6477.72***
Total monthly expenditure (Tk)	16,365.49	475.31	14,167.65	425.05	−2197.85***
Shortage of quality seed (Yes/No)	0.633	0.483	0.738	0.441	0.104***
Shortage of pesticides, insecticides/herbicides (Yes/No)	0.661	0.474	0.756	0.430	0.094***
Lack of training facilities (Yes/No)	0.787	0.410	0.816	0.388	0.029
Shortage of labor in production (Yes/No)	0.614	0.487	0.650	0.478	0.036
Higher input price (Yes/No)	0.774	0.419	0.906	0.292	0.132***
Transportation problem (Yes/No)	0.091	0.288	0.575	0.495	0.484***
Labour shortage in output (Yes/No)	0.097	0.297	0.266	0.442	0.168***
Shortage in demand (Yes/No)	0.163	0.370	0.653	0.477	0.490***
Inadequate information (Yes/No)	0.235	0.425	0.669	0.471	0.434***
Problem in fair pricing (Yes/No)	0.179	0.384	0.678	0.468	0.50***

Note: *** p<0.01, ** p<0.05, * p<0.1

4.2. Problems experienced by farmers during the COVID-19 pandemic

Farmers in Bangladesh were experiencing a variety of issues as a result of the COVID-19 outbreak in the initial time period (2020). Those problem items were gathered from various research papers, print and electronic news media outlets, as well as in-depth conversations with key experts.

In the case of input market-related problems faced by the farmers, a total of seven problem items were identified and 320 people replied to each problem item; the PCI varied from 0 to 594, with “0” indicating no problem and “594” indicating the most severe problem. The descending order of PCI was also used to build the rank order. The first part of the following table (Table 3) shows how the input-related problem items were rearranged based on rank order.

Table 3. Rank order of input and output market-related problems faced by farmers during COVID-19

Item no.	Input related problem items	Faced severe problem (3)	Faced moderate problem (2)	Faced low problem (1)	Faced no problem (0)	PCI	Rank order
7	Higher input prices	102	100	88	30	594	1
3	Lack of technical knowhow	74	84	103	59	493	2
1	Shortage of quality seeds	52	69	115	84	409	3
2	Shortage of insecticides/herbicides	45	73	124	78	405	4
6	Scarcity of fertilizers	38	71	121	90	377	5
5	Shortage of labour	23	59	126	112	313	6
4	Inadequate irrigation	13	19	153	135	230	7
	Output related problem items
6	Lack in fair pricing	70	67	58	125	402	1
5	Inadequate information	55	73	86	106	397	2
4	Shortage in demand	50	75	88	107	388	3
3	Labour shortage	52	48	88	132	340	4
1	Transportation problem	33	71	87	129	328	5
2	Packaging problem	18	11	3	288	79	6

Source: Analysis of authors’ survey.

The findings revealed that, among the seven problem items, the majority of the farmers (32%) identified “higher input prices” as a severe problem during the COVID-19 pandemic period (2020). This was because during the pandemic there were movement restrictions within the nation and also at the borders, which ultimately interrupted the smooth supply of inputs in all places, or the supply was delayed and due to the shortage of supply, farmers had to pay higher prices for inputs. The “lack of technical training facilities” was the second most serious issue, as there were frequent lockdowns and social distance was maintained for fear of infection. Moreover, the third most severe problem was a “shortage of quality seeds”. The least concerning problem was identified as related to the irrigation facility because the lockdown and restrictions were not directly related to the irrigation system or its operation.

In the case of output markets, a total of six problems were identified and 320 farmers put their scores on each problem. These are the specific problems that were faced by the farmers while they attempted to sell their vegetables after harvesting in the local and regional markets due the pandemic issue.

Table 3 also shows the rank order of the problems according to the PCI, related to the output market of vegetables faced by the surveyed farmers in the study area. The PCI varied from 0 to 402, with “0” indicating no problem and “402” indicating the most severe problem. Among the six problems, “lack of fair pricing” (61%), “inadequate information” (67%), and “shortage of demand” (67%) were identified as the top three problems by the majority of farmers. According to farmer responses, labor shortage (59%) during harvesting and marketing of vegetables, and transportation problem (60%) were the types of moderate issues they faced during the pandemic year.

Initially, the movement restrictions stopped the transport of agricultural products, which cut farmers off from their customers. Also, the lack of traders at the farm gate or local market (middlemen like collectors, transporters, wholesalers, etc.) and vehicles kept farmers from getting the fair market price.

The survey results (Table 4) reported that the average vegetable production during the pandemic (2020) was lessened by 6.18% compared to the previous year (2019). Consequently, the average sale of vegetables decreased by 3%. Due to the reduction in sales, the quantity of marketed surplus decreased and marketable surpluses increased in farm households. Vegetables are highly perishable, so farm households were not able to store their produce for further sale. They had to utilize a huge quantity of their products by consuming more and by distributing to others (relatives, neighbors, friends, etc.) or feeding their livestock.

Table 4. Changes in farm output during COVID-19 pandemic

Quantity (kg)	Before Covid (2019)	During Covid (2020)	Mean Difference (2020–2019)	t-value
Average production	13,578	12,739	−1,087	−1.16
Average consumption	464.63	466.15	1.52	0.04
Average distribution (gift and others)	394.06	640.93	246.87***	3.34
Average sale	12,719	11,632	−838	−0.86

Note: *** p<0.01, ** p<0.05, * p<0.1.

4.3. Impact of COVID-19 pandemic on market access of vegetable farmers

The result from the impact assessment analysis through a probit regression model is shown in Table 5. The model indicates the association between the dependent variable (percentage of marketed surplus or sales of vegetables), which is a proxy indicator of the market access ability of the farm households, and the treatment variable (time of year). The model represents the coefficient values of all the demographic characteristics, different problems related to the input and output markets of vegetables and region dummies, which are the explanatory variables of this study and their association with the treatment variable. The result shows that households with older farmers were affected significantly during the pandemic. The total household income was significantly reduced and the total household expenditure was significantly increased during the pandemic.

Table 5. Determinants of the changes in marketed surplus

Variables	Co-efficient	Standard error	Marginal effects
Gender (1=Male)	−0.138	0.244	−0.055
Ln Age	0.457**	0.233	0.182
Marital status (1=Married)	0.230	0.234	0.092
Family member (No.)	0.013	0.037	0.005
Primary	0.119	0.182	0.047
Secondary	0.128	0.195	0.051
Higher secondary and upper	0.255	0.232	0.101
Ln Total monthly income	−0.647***	0.095	−0.257
Ln Total monthly expenditure	0.451***	0.091	0.179
Market distance (km)	−0.039	0.040	−0.015
Shortage of quality seed (1=Yes)	0.310*	0.183	0.123
Shortage of pesticides, insecticides/herbicides (1=Yes)	−0.147	0.206	−0.058
Lack of training facilities (1=Yes)	0.066	0.214	0.026
Labor shortage in production (1=Yes)	−0.309*	0.167	−0.123
Higher input price (1=Yes)	0.744***	0.236	0.295
Transportation problem (1=Yes)	1.304***	0.276	0.518
Labour shortage in marketing (1=Yes)	−1.022***	0.259	−0.406
Shortage in demand (1=Yes)	0.507***	0.181	0.201
Inadequate information (1=Yes)	−0.138	0.219	−0.055
Problem in fair pricing (1=Yes)	0.834***	0.254	0.331
Dinajpur	−0.257	0.193	−0.102
Rangpur	−0.430**	0.196	−0.171
Pabna	−0.137	0.196	−0.055
Constant	−0.814	1.268	-
LR Chi square	323.95***
Pseudo R^2	0.366
Log likelihood	−280.95
Observations	639

Note: *** p<0.01, ** p<0.05, * p<0.1.

The pandemic occurred at a time when the farmers were preparing to sell their Rabi crops after harvest, and the lockdown continued during the production time of Kharif-1 crops. As a result, the farmers faced problems both in marketing the harvested products and acquiring production inputs for the next season. The model shows that all of the problems with the input and output markets for vegetables and the effects of a pandemic are linked and depend on each other. Among all the identified problems, a lack of quality seed, higher input prices, transportation issues, a lack of demand, and a lack of fair pricing are found to be positive and significant determinants of the COVID-19 pandemic’s impact. Farmers in the Rangpur district were significantly more affected by the pandemic than farmers from other districts. Because agriculture is the main source of livelihood for the majority of the population in Rangpur, and the agricultural sector is characterized by small and fragmented farms, low productivity, and limited access to markets. Besides, it is considered as one of the poorest regions in the country, as the poverty rate was around 41% in 2016, which is higher than the national average of 31% (HIES, 2016). However, the marginal effect analysis indicates that one unit change in each significant problem has increased the probability of being affected by 12% to 52% during the pandemic. With the change in income and expenditure per unit, the probability of being affected went down by 26% and increased by 18%, respectively. In other words, individuals with higher total monthly income are less likely to be affected by the pandemic compared to those with lower total monthly income and again individuals with higher total monthly expenditure are more likely to be affected by the pandemic compared to those with lower total monthly expenditure.

The results from Table 6 represent the impact of COVID-19 on the marketed surplus of vegetables. Based on different matching techniques, the ATT values are found on the outcome variable. The ATT values are almost similar in all the matching methods, confirming the robustness of the findings. The impact of the COVID-19 pandemic on the marketed surplus of vegetables was found to be around 3–5%, meaning that the marketed surplus or sale of vegetables was significantly decreased by 3–5% on an average during the pandemic time period (2020) due to the impact of the determinants related to the COVID-19 pandemic.

Table 6. Impact of COVID-19 on marketed surplus

Matching	ATT values	Standard error
Nearest Neighbor Matching	−3.193***	0.705
Radius Matching	−4.474***	0.738
Kernel Matching	−4.439***	1.598
Stratification Matching	−4.973***	1.402

Note: *** p<0.01, ** p<0.05, * p<0.1.

4.4. Coping strategies followed by farmers during the COVID-19 pandemic

COVID-19 has put farmers’ livelihoods at risk. The surveyed farmers were significantly reliant on vegetable income to survive and continue farming, with approximately 38% of their income coming from vegetable production. Though COVID-19 affects people in various ways, the most major one is falling income. Among the 320 surveyed farmers 79% reported that there was reduction in their family income during the COVID-19 pandemic.

Table 7 shows different strategies that farm households followed to cope with their income loss during the pandemic. The income shock effect influenced them to adjust immediately, mainly through selling their assets, reducing household expenditure and borrowing credit from different sources. Among all types of strategies, borrowing credit was found to be the most popular among the respondent households. On the other hand, selling valuable assets is not so easy for the typical farm households, though it was found that the second highest group of households sold their livestock resources for cash. The households also tried to minimize their income loss by cutting down the household expenditures on food and non-food items (such as education) (Gatto & Islam, 2021).

Table 7. Different strategies followed by the farm households to cope with the COVID-19

No.	Strategies	Frequency	Percent
A	Selling household assets
1	Selling land	10	3.13
2	Selling livestock resources (cattle, goat etc.)	130	40.63
3	Selling gold jewelers	4	1.25
B	Reducing household expenditure
1	Food expenditure	88	27.50
2	Non-food expenditure	87	27.19
C	Borrowing credit from formal or informal sources	138	43.13
D	Others	50	15.63

Table 8 shows that the mean difference for all income variables is negative, indicating the significant income decline in 2020. For the expenditure variables, the mean difference is negative for total family expenditure, education expenditure, and food consumption expenditure, indicating that these expenditures decreased in 2020 compared to 2019. The downturn in farm household`s income led to a reduction in their overall expenditures, with the exception of the cost of medical care, which increased due to the severe health risk caused by the pandemic. On the other hand, due to the pandemic’s effect on income, the average amount spent on education also dropped considerably.

Table 8. Changes in household income and expenditure due to COVID-19 pandemic

Items	Mean difference (2020–2019)	t-value
Income (Yearly in Tk.^1)
Total family income	−76,879***	−4.56
Total farm income	−57,176***	−3.91
Farm income from crops	−7,153	−1.57
Farm income from vegetables	−38,362***	−4.99
Farm income from fruits	−2,132	−0.32
Farm income from livestock	−8,490	−1.60
Farm income from fisheries	−1037.50	−0.98
Total off-farm income	−19702.66***	−2.68
Expenditures (Yearly in Tk.)
Total family expenditure	−2,169.63***	−3.41
Food consumption expenditure	−423.44	−1.15
Education expenditure	−2,089.69***	−8.08
Medical expenditure	390.63*	1.86
Utility expenditure	58.025	0.84

Note: *** p<0.01, ** p<0.05, * p<0.1

Generally, poor households used to cope with financial shocks by eating less nutritious foods and more grains and relying on public food relief programs for survival (Ho¨hler and Lansink, 2020). Such practices can make them prone to health risk due to malnutrition which may ultimately lead to lost employment or lives and a further cycle of poverty. Farmers also expect to experience difficulty in acquiring inputs owing to income loss as they also need cash urgently to smoothly commence their future production.

From the survey data (Figure 1), we found that during the COVID-19 pandemic, households experienced a change in their food consumption habits by either cutting down or increasing the types of food items in their basket. About 38% of farm households reported that they had increased consumption of vegetables during the pandemic year (2020). This is a clear indication that the households had to consume their own produced vegetables due to the inability to sell them on the market. On the other hand, around 74% of households had to decrease their consumption of meat (poultry, beef, mutton, etc.) and around 62% of households had to cut down their consumption of fish (small and big) during that time, which means the households didn’t have enough income or cash to spend out on high-value nutritious foods due to the price spikes or might face rumors about food safety concerns (Alam et al., 2021; Sunny et al., 2021; Zabir et al., 2021). However, some households (14%) shared that the notion of drinking more tea and spices to boost human immunity against COVID-19 also influenced them during the pandemic.

Figure 1. Percent change in consumption of different food groups.

Vegetable farmers encountered many challenges obtaining inputs for production as a result of the pandemic. During the pandemic, farmers reported having to obtain input supplies from various sources whenever they faced higher input prices, a shortage of quality seeds, a shortage of fertilizers, insecticides, and herbicides, a shortage of labor, and so on.

Most farmers used to rely on government authorized dealers and non-government local sources for fertilizer. During the pandemic, the number of these farmers slightly declined, and they said they had to use compost instead of chemical fertilizers more often. On the other hand, vegetable farmers mostly depend on the government-authorized dealers for quality seeds, and despite the problems due to the pandemic, the proportion remained unchanged. In the case of pesticides, herbicides and growth hormones, the proportion of farmers dependent on non-government local distributors declined slightly towards the government source (Figure 2).

Figure 2. Percent of farm households collected inputs from different sources before and during COVID-19.

During the pandemic year, a shortage in labour force has been noted as a prevalent issue in the production and harvesting times. Initially, the workforce was unavailable, and the labour expected a much higher wage (Khan et al., 2021). Farmers typically employ both family and hired labour in their farming activities. As a result of the pandemic, female labour participation grew more than male labour participation. In comparison to the previous year (2019), the share of males in the family labour force has increased substantially. On the other hand, the number of females hired workers dropped less than the number of males hired workers (Figure 3).

Figure 3. Percent of labour participation before and during COVID-19 .

Approximately 34% of households questioned had obtained credit from formal and informal sources to alleviate their financial difficulties during the pandemic year (Table 9). The majority had borrowed loans from non-governmental organizations (NGOs) and then from informal sources. Roughly 7% of the borrowers claimed that they could obtain loans from banks with a 4% interest rate. In contrast, the majority of borrowers who got loans from non-governmental organizations will be required to pay interest rates that exceed 4%. Some farmers also reported that NGOs provided them with loans at an interest rate of less than 4%. However, very few farmers could obtain loans from informal sources with an interest rate lower than 4%.

Table 9. Credit taken from different sources and percentage of interest to the farmers

Sources	Credit taken (%)	Less than 4%	Equal to 4%	More than 4%
Bank	4.38	0	6.86	13.73
NGO	20.63	4.90	0.98	59.80
Microcredit institution	0.31	2	0	5.31
Friend	1.24
Relative	4.38
Businessman	0.31
Farmers’ cooperative	1.88
Others	0.63

During the COVID-19 pandemic, the government of Bangladesh tried to maintain motivation of farmers to utilize all types of land and encouraged the agricultural extension officers to stay at sites to support farmers. Extension services tried to broaden their activities to cover the COVID-19 situation and properly support farmers by providing information. Farmers also proactively tried to contact them to receive support and services (Figure 4). Among the interviewed farmers nearly 65% said that they have acquired extension services and out of those, 62% farmers got their required services directly by meeting with the field officers. Telecommunication services also privileged one-fourth of those farmers to reach the extension personnel whenever needed.

Figure 4. Percent of farmers received extension services during the COVID-19 period.

4.5. Government initiatives to mitigate the impact of the COVID-19 pandemic

Despite the fact that Bangladesh may have had a crisis in the food system in terms of production and distribution, the government took a number of steps to combat this shock. In response to COVID-19 related lockdowns, the government of Bangladesh distributed rice, potatoes, lentils, oil, onions, and salt to individuals in need around the country. It also regulated the sale of rice on the open market at Tk10 per kg in the country’s capital and major cities. In view of the COVID-19 aftermath, the Prime Minister of Bangladesh proposed a Tk 5,000 crore stimulus program for farmers to support agricultural production. This revolving refinance program was announced for farmers in the agriculture, dairy, poultry, horticulture, and fisheries sectors. Variable-amount subsidies have been granted to farmers to enable them to acquire agro-machinery to cut production costs and boost crop yields. The government launched an emergency transportation system for agricultural products but a significant amount of perishable items continued to be lost every day due to the absence of a proper distribution and monitoring system.

From the respondents of this survey (Figure 5a), it has been reported that, among the 320 farmers, only 8.13% had received the emergency donation allowance properly from the government. Only 1.56% of farmers reported receiving agricultural machinery for irrigation from the government, and less than 1% received emergency transportation to distribute their products in the supply chain’s long-distance markets. The emergency transportation system for agriculture could not support the vegetable farmers properly and a significant amount of perishable items were being wasted every day as a result of an inadequate distribution and monitoring system (Zabir at al., 2020).

Figure 5. a) Percent of farm households received government allowances during the COVID-19 and b) Different social safety nets accessed by participants.

Moreover, 18.75% farmers reported that they had received different benefits from different government allowance programs (Figure 5b). Among those beneficiaries, almost 35% of farmers reported that they had got the facility of open market sale during the pandemic. Despite the pandemic, beneficiaries have also received the regular social safety net programs, such as the old-age allowance, widower’s allowance, VGF card, food for work, and others.

5. Conclusion and Recommendations

COVID-19 affected agriculture on a global scale, but its impacts were most severe in regions already impacted by climate change and poverty. This pandemic created major negative consequences for perishable products and their supply chains. Perishable goods require quick distribution, marketing, and export. However, the outbreak of the pandemic led to the deterioration and distortion of many of these products as a result of the social distance restrictions imposed by governments to save the lives of the people and the transportation shutdowns that made distribution impossible on both domestic and international markets. Farm households in developing countries like Bangladesh have faced problems of different dimensions. The COVID-19 situation was highly unpredictable, and vegetable farmers reported major disruptions in their production and marketing activities, a reduction in household income, less consumption of diversified nutritious food, the sale of household assets to manage cash flow, and credit taken to mitigate the financial difficulties, all of which have consequences in the long term.

The results of this study indicated that farmers faced input market constraints including high input prices, a lack of technical training, and a lack of quality seeds during the pandemic. For output markets, the major problems were identified as a lack of fair pricing, inadequate information, and a shortage of demand during the COVID-19. Moreover, the participation of hired labor decreased due to the problem of labor shortage, and as a counter measure, the participation of family labor, especially the ratio of female labor, increased within farm households. The vegetable farmer had lost almost 3–5% of their marketed surplus, which could not be sold in the market due to the disruptions raised by this pandemic. Complex risk-coping behaviors have been developed by farming households, such as sourcing out farm inputs, reducing food intake, selling properties and assets, obtaining credit and loans from formal and informal sources, taking cash or non-cash kinds, and increasing work. The farm households had experienced a change in their food consumption habits during the pandemic, where the consumption of animal protein decreased significantly. Most of the farm households borrowed credits from NGOs with an interest rate of above 4% to cope with their financial crisis. Although agricultural stimulus packages have been offered by the government, they did not provide clear incentives for smallholder vegetable producers who have faced significant losses on their crops and were not equitably distributed or accessed. The proportion of farmers and their households who received cash incentives from the government was not considerable, but they had the facility of open market sales during the pandemic. Despite the pandemic, beneficiaries have also received the regular social safety net programs, such as the old-age allowance, widower’s allowance, VGF card, food for work, and others.

Global health emergencies will continue to occur as long as plants, animals, and humans cohabit on earth, resulting in substantial health and economic losses. Over the next few decades, both the global population and rates of urbanization will increase, pandemics will become more widespread, and climate change will worsen. As a result, resilient food systems must be upheld as our society transitions to sustainable development and a climate-neutral economy. Thus, rather than a post-pandemic return to development as usual, the government should take initiatives that advocate for a shift to sustainable, resilient, and inclusive development. In light of the findings of this study, it can be recommended that the collaborative participation of the relevant sectors is needed to decrease the detrimental effects of a shock like COVID-19 on perishable items by employing relevant mechanisms. Moreover, building farmers’ understanding of the crisis or shocks, adaptability, mechanization and use of digital technologies is crucial to making considerable changes for future disruptions. The recovery from the pandemic presents the opportunity for broader promotion and adoption of proven innovations to improve livelihoods with enhanced natural resources, such as diversifying farming systems to increase productivity and food security. More importantly, it is crucial to set up dedicated vegetable markets, provide improved transportation and storage facilities, ensure fair prices for the farmers, and implement training and capacity-building programs to improve their production and marketing skills. Furthermore, the government should take policy measures to create a national emergency fund to provide financial support to all farmers with proper monitoring and distribution systems when emergency situations arise in the future.

6. Limitation of Study

The main limitation of our study is related to the time constraints. We conducted our field survey from April to May 2022 and some information were collected by recalling from farmers. During and after the COVID-19 lockdowns, there were some restrictions including in travelling, conducting face-to-face interviews and vaccine compliance issues, and it was very difficult to collect farmers data through an online platform. As a result, we had to wait to interview them in person. Although we tried to complete the survey in a suitable time period when the reliability of data could be achieved. In Bangladesh, from the mid of April, Bengali New Year starts, so farmers usually calculate their earning and spending before the New Year and planned for the next year. So, we utilized this time period to complete our survey. Another limitation of our study is related to the sample size as we only captured 40 randomly selected farmers from each upazila due to the budget constraints.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

The work was supported by the IFPRI and ACIAR .

Notes on contributors

Mohammad Jahangir Alam

Mohammad Jahangir Alam is Professor at Department of Agribusiness and Marketing, Bangladesh Agricultural University. He was a Visiting Fellow at Australian National University. He was a Fulbright Scholar at Cornell University, USA and was a Commonwealth Fellow at SOAS, UK.

A.N. M Faijul Kabir

A. K. M. Faijul Kabir is currently an Associate Scientist at the IRRI Bangladesh. He was a Research Assistant at the Department of Agribusiness and Marketing, Bangladesh Agricultural University.

Tamanna Mastura

Tamanna Mastura is currently a Research Assistant at the Department of Agribusiness and Marketing, Bangladesh Agricultural University.

Avinash Kishore

Dr. Avinash Kishore is a Research Fellow in the New Delhi Office of the IFPRI. He has a Ph.D. in Public Policy from Harvard University and a Master’s in Public and International Affairs from Princeton University.

Tamara Jackson

Dr Tamara Jackson has worked on farming and food systems projects that spanned a range of goals from improving farming and marketing systems and stakeholder engagement to foster communication and collaboration.

Ismat Ara Begum

Dr. Ismat Ara Begum is a Professor in Agricultural Economics at Bangladesh Agricultural University. he holds a PhD and a M.Sc in Agricultural Development Economics from Hokkaido University, Japan.

Notes

1. 1$ = 106.32 Tk. (2023).