To conduct the study, five students’ hostels, as shown in Figure 2, were selected as the study areas. Female household workers are the target group for this study. In Rajshahi city, their workplaces are divided into three categories: houses, student hostels, and both.

A preliminary survey was conducted to identify the HH workers. This study was conducted in 2022, after the COVID-19 pandemic. Students’ hostels were selected because the HH workers working in hostels also work in houses. Several HH workers who were working in hostels at the time of this study were employed in houses before the COVID-19 pandemic. This study tried to present the conditions of the target group in all workplaces before and during the COVID-19 pandemic period.

A total of 30 female workers were found in the hostels and selected as the sample to carry out the study.

Figure 2. Map of the study area (circles show the study areas).

A questionnaire survey was conducted to collect data on demographic characteristics, the workplace of household workers, changes in income level, income source, and workplace during COVID-19, and so on. These data denote the livelihood and income structure of the target group.

Furthermore, data on the food intake structure of the workers’ households were collected. Data on food consumption and dietary intake of the households of the workers’ families were collected. Also, food collection processes by the workers to manage their households were identified.

Finally, the coping strategies to address the impact of COVID-19 on food security were identified.

Data collected from the questionnaire survey were analyzed using Microsoft Excel. Basic demographic data, livelihood status, and income data have been analyzed and represented in bar charts, pie charts, and tabular format.

Food patterns were analyzed using the Food Consumption Score (FCS) and the Household Dietary Diversity Score (HDDS).

Data on food consumption and food diversity have been calculated following the food charts and equations below:

Table 1. Food chart for calculating FCS.

Source: [21].

Following this chart and considering the weight of each food group, the equation below is applied to calculate FCS.

where,

xi = Frequencies of food consumption = number of days for which each

ai = Weight of each food group

*Food group was consumed during the past 7 days (7 days were designated as the maximum value of the sum of the frequencies of the different food items belonging to the same food group)

FCS is assessed by considering the following ranges:

Poor consumption = 0 - 21

Borderline consumption = 21.5 - 35

Acceptable consumption ≥ 35

After that, the Dietary diversity of the household has been obtained through calculating the Household Dietary Diversity Score (HDDS) with the following food chart:

Table 2. Food chart for calculating HDDS.

Source: [22].

*Food group was consumed during the last 24 hours is the consideration and the information was collected through questionnaire.

The following procedure is followed for calculating HDDS:

1. Regrouping the 16 food groups used for FCS into the 7 food groups.

2. For each food group, a new binomial variable that has two possible values.

1 - yes: the household/individual consumed that specific food group.

0 - no: they did not consume that food.

3. Sum all the binomial variables to create an HDDS.

The threshold values for HDDS are:

6+: high = good dietary diversity

4.5 - 6: medium dietary diversity

<4.5: low dietary diversity

Besides, food sources have been presented showing the percentage in a pie diagram. The link between the income of household workers and the FCS and HDDS has been established both for the COVID-19 pandemic and before that.

Food costs for each household were calculated after the preparation of a food chart for seven days a week. The information on coping strategies adopted by female workers during the pandemic was collected from a questionnaire and prioritized through a rank scoring method. The strategy with the highest score was ranked 1 and considered a top priority. Likewise, other coping strategies were ranked and prioritized. The result was shown in a stacked column.

The study shows that the age group of most (45%) of the female household workers is in the 30 - 40 category (Figure 3). The study found that nearly 50% of the household workers’ education level was Class 1 to 5 (Figure 4).

In the study, the livelihood patterns of female household workers in the selected study areas have been analyzed from two perspectives: workplace and income. In the study areas, it is found that before the COVID pandemic, 48% of workers

Figure 3. Age distribution of the selected female household workers.

Figure 4. Level of education of the selected female household workers.

worked in houses. The number of workers working in hostels is relatively low. Besides, the percentage working in both places was 31%, not negligible.

Analysis of the workplace of workers revealed different categories during the COVID-19 pandemic, particularly during the lockdown period. The categories include having no job or being fired from work; occasionally working in houses or hostels; occasionally working in both places; regularly working in any of the places; and working in the same places before the COVID-19 period. Table 3 shows that a maximum (28%) of workers worked either in houses or hostels occasionally, in case of any religious occasion, wedding ceremony, or party, etc. Additionally, a similar percentage (27%) of workers were fired from their jobs due to concerns about the safety of their households.

Table 3. Working places of the female HH workers before and during COVID-19.

In this study, the income categories of female household workers have been categorized as follows: BDT less than 5000, BDT 5000 to 8000, BDT 8000 to 11,000. Before the pandemic, the average monthly income of a household worker was BDT 5000 to 8000 (52%). Also, the BDT 8001 - 11,000 category was visible enough (38%).

However, the scenario changed during the COVID-19 lockdown. Around 45% of the workers had no income, and 35% had an income of less than BDT 5000, whereas before COVID-19, the percentage was only 10% (Table 4).

Table 4. Percentage of monthly income of the female household workers of the study area during and before the COVID-19 periods.

One of the concerns of this study is to observe the food pattern during COVID-19, and a comparison with the period before COVID-19 of the households of the female workers. Food Consumption Score (FCS) denotes the food consumption pattern of the households. It provides the intake of food within the last seven days. From the study, it is identified that the food consumption pattern of the households is not poor (range 0 - 21), both pre and during COVID-19 conditions. At the pre-stage, the level was at the borderline (range: 21 - 35), and in most cases it is acceptable (range: >35). When it comes to observing the food consumption condition during COVID-19, the scenario says that the proportion of acceptance level in the case of food consumption was less than before the pandemic (Figure 5).

Figure 5. Variation in Food Consumption Score (FCS) before and during the COVID-19 period.

On the other hand, the diet of the households was not diversified (Figure 6). Both pre and during COVID-19 periods, the households did not consume diversified food. Thus, the level “Good diversity (>6)” is blank in Figure 6. In the COVID-19 pandemic, the household’s food diversification score is very poor (range <4.5). Whereas before COVID-19, they had a medium (4.5 - 6) level of food diversity.

Figure 6. Variation in household dietary diversity score (HDDS) before and during COVID-19.

From the field survey, a generic food chart for one person of the family for a week was generated (Table 5). They mainly consume rice or carbohydrates. Alternatively, pulses and vegetables are present almost every day. Here, rice or potato or bread is considered the main staple. Sometimes, or very few times, they intake fruits or milk. Also, eggs and green vegetables were included in the food list during COVID-19. They do not starve, but the variation is below standard.

Table 5. A generic food chart of a household.

According to this food intake pattern, the average per day per person minimum food cost is approximately BDT 80 - 100. The family structure of an HH worker is shown in Figure 7. Most families have 4 to 5 members. Considering the per-day cost of food for an adult person, the monthly cost for each household is a minimum of BDT 160 to 500. Total food cost per month was approximately BDT 13,000 - 15,000.

Figure 7. Percentage of family members of the female household workers.

Sometimes, the items vary due to special occasions. Most of the HH workers have other earning members in their family (Figure 8). In this case, with this financial support, food costs were bearable before the COVID-19 period. However, during the pandemic, bearing the high food costs was troublesome for the workers. Another important issue here is the sources of food during the two periods. Most of the time, female household workers avail food from their workplace (house/hostels/both). Also, they buy food from the market. Nearly 50% of female workers get food from both their workplace and the market (Table 6). The food they get from their workplace is enough for three meals a day. It is also worth mentioning that, in most cases, female workers work more than one place in a day.

Figure 8. Presence of other earning family members.

During the COVID-19 pandemic, they had to spend money on buying food from the market (45%) (Table 6). Very few female workers were still in their jobs during the pandemic and got three times the food. Non-Government Organizations (NGOs) such as Quantum Foundation, BRAC, JAGORONI Foundation, etc., provided aid in some areas. Initiatives from the Trading Corporation of Bangladesh (TCB) were undertaken to distribute food at a cheap rate. However, due to mismanagement, the support did not reach the poor.

Table 6. Food collection source before and during COVID-19.

Figure 9 and Figure 10 illustrate the changes in food consumption and diversification patterns in response to income variation during both pre- and during COVID-19 periods. During the lockdown, food consumption was at the borderline as the income was less than 5000 BDT. Across all income ranges, COVID-19 led to lower food consumption in households than before COVID-19 periods (Figure 9). In the case of dietary diversity, the pattern is also not good within a similar income range (below BDT 5000) (Figure 10). An interesting observation from this analysis is that at an income level less than BDT 5000, the FCS is acceptable before COVID-19 period, but the HDDS is poor at the same time.

Figure 9. Income-wise FCS variation before and during the COVID-19 periods.

Different strategies include cutting a meal one day; spending the income to feed all the other members of the family three times a day and consuming less on one’s own; prioritizing the male and kids; ensuring one’s own meal first; starving and providing the available food to all the members, and well distribution of food after managing from different sources.

Figure 10. Income-wise HDDS variation before and during the COVID-19 period.

According to the workers, the distribution of food among all the members should be as much as possible. The second strategy they put in place is to spend their income to ensure food three times a day. After that, they focused on providing the food to the kids first. Some of the workers have small children. The female workers then fixed their target to feed the children first. Chronologically, the other strategies include continuing to starve and distributing available food to others. Occasionally, they prioritize male members: husbands, sons, brothers, and others. Lastly, they emphasized cutting out meals for the day. Figure 11 shows the prioritization of different coping mechanisms.

Figure 11. Prioritization of the coping strategies during COVID-19 to manage food security.