Introduction

Background

Globally, waste generation is increasing, with cities presently producing more than 1.3 billion tons of waste annually and increasing to 2.2 billion tons by 2025 [1]. The total of 33% of waste produced everyday remain uncollected, and around 3.5 billion people lack basic waste disposal services [2]. Improper waste disposal is associated with the spread of diverse diseases [3]. Slum households represent the worst case of waste disposal in urban areas. Slum dwellers suffer from diseases resulting from air, soil and odor pollution caused by waste dum** in nearby areas. Slum dwellers suffer from seasonal and waterborne diseases due to inadequate solid waste disposal, drinking water, and sanitation [4]. Approximately 2.23 million people living in slum areas of Bangladesh suffer from disease related to the mismanagement of waste [5].

Waste disposal facilities in urban centers are insufficient for people living in slums. Financing waste disposal through a top-down approach, such as public expenditure, is not effective enough to ensure better waste disposal, so a bottom-up approach needs to be implemented to ensure better waste disposal. Slum dwellers have a lower monthly income, a lower level of education, and less access to waste disposal services than do the general population, making them unwilling to spend money on waste disposal [6]. Lower expenditures on waste disposal signify less willingness to manage waste efficiently [5]. Waste disposal expenditures include the expenditures incurred to collect, sort, deliver and dispose of waste [7]. Therefore, increasing household-level expenditures in different contexts of waste disposal, such as eco-friendly practices, hygiene, and slum sanitation systems, is critical in urban centers.

The lifeline of remittance development for develo** countries reached 529 billion USD in 2018 globally [8]. Furthermore, remittances are associated with household-level welfare Nanziri et al. [9]. When a slum household receives remittances, the household must decide whether the income from remittances should be spent on basic needs such as food, clothing and housing or on health needs, such as waste disposal. Latif et al. [4] estimated that slum dwellers spend 61.39% of their income on food. Most of their resources are spent on activities other than waste disposal. As such, there is a need to devise ways to enhance and increase the expenditure of sum households on waste disposal to improve their health. One such way includes the use of remittances.

No previous research has focused on the financial mechanism of slum household waste disposal, and this study attempts to address this research gap. This study examines the nexus among remittance receipts, waste disposal and health outcomes. Two research questions are addressed in the study. Firstly, this study examines the impact of remittances on total monthly waste disposal expenditures. Secondly, this study examines the impact of waste disposal on the total monthly health expenditure of slum households in Bangladesh. Our findings have implications for how remittances influence the waste disposal behavior of slum dwellers. Expenditure on waste disposal suggests a willingness to enhance quality of life. Total health expenditure is considered an indicator of the health conditions of slum households.

Determining the causal impact of remittances on monthly waste disposal expenditures and the subsequent influence of improved waste disposal on monthly outpatient expenditures is challenging due to the nonrandom assignment of remittances and waste disposal decisions. The presence of self-selection bias, where treated units choose to receive the treatment, introduces bias to the results. This study uses propensity score matching (PSM) to address the selection bias that occurs because of comparisons between treated and control units. Propensity scores serve as a helpful method for accounting for observed covariates in observational studies focused on determining causal effects [10].

Literature review

To the best of our knowledge, no previous studies have focused on the relationships among remittance receipts, waste disposal and the health conditions of vulnerable households. Many studies have concentrated on these three concepts: remittance receipt, waste disposal and health. A number of studies have focused on willingness to pay for waste disposal services. These studies provide insight into how much households are willing to pay for disposing of their waste. Afroz et al. [11] calculated the willingness to pay for waste disposal, arguing that if every household contributes approximately 13 Bangladeshi Taka (BDT) per month for waste disposal services, it could generate a monthly sum of 7.6 million BDT in Dhaka city. Furthermore, Han et al. [7] reported that male and wealthy individuals have greater willingness to pay than do female and poor individuals. Similarly, Kayamo [12] argued that local governments should consider household-level willingness to pay for enhanced solid waste management services and endorse projects aimed at improving solid waste management to mitigate potential public health hazards and environmental consequences. Moreover, Suryawan and Lee [13] argue that rich households are more willing to pay for adaptive waste management services than poor households are.

Very few studies have examined financial resources related to waste disposal. Alhassan et al. [14] reported that monetary incentives can encourage better waste disposal at the household level. Pinha and Sagawa [15] argued that a well-structured financial mechanism that includes the cost and revenue of solid waste disposal can increase the effectiveness of waste disposal programs. Furthermore, Welivita et al. [16] examined different charging methods for solid waste disposal mechanisms and suggested that the prepaid bag-based charging method is more effective in develo** countries. Few studies have explored interventions to improve household-level waste disposal. Immurana [17] examined financial inclusion as a financing mechanism to improve solid waste management. However, no previous studies have examined any financial mechanism that enables households to spend on solid waste disposal. Few studies have connected remittance with sanitation practices and spending patterns at the household level. In addition, Tsafack and Djeunankan [18] argued that remittances increase access to better sanitation services at the household level. Moreover, Raihan et al. [19] argue that remittances enhance household-level expenditures on health. Similarly, Hua et al. [20] argue that remittances increase expenditures on household repairs. However, no research has examined the impact of remittances on waste disposal spending as a financial mechanism.

Furthermore, some studies have attempted to determine the correlation between waste mismanagement and health outcomes. Adeniran and Shakantu [21] explored the linkages among waste disposal, health and environmental issues, claiming that waste disposal dumpsites are negatively related to health outcomes. Similarly, Siddiqua et al. [22] explored correlations among environmental pollution, health and waste landfilling. Furthermore, Vinti et al. [23] examined context-specific practices of waste disposal and exposure pathways that result in health risks. Moreover, Sangkham [24] explored the impact of medical waste disposal on the health crisis during the COVID-19 pandemic.

Under the backdrop of these previous studies, two research gaps are addressed in the present study. First, no previous studies have examined the impact of remittances on waste disposal expenditures. Waste disposal expenditure is an indicator of both the willingness and capability of a household to dispose of waste efficiently through third-party waste disposal companies. This study explores the causal impact of remittances as a micro level financing mechanism on waste disposal expenditures. Second, this study examines the impact of better waste disposal on health outcomes. The impact of waste disposal on health outcomes will provide empirical evidence on the significance of managing waste efficiently at the household level.

Materials and methods

Data description and covariate selection

Only the slum households included in the Household Income and Expenditure Survey (HIES) 2016–2017 are considered. Random sampling was employed for the HIES, so we considered the data from the slum households selected as the sample to represent the overall data of all the slum households in Bangladesh. This study included 1773 urban households as the units of analysis. First, when estimating the impact of remittance receipt, the treated households received remittances, while the control households did not. A total of 178 households in the treated group received remittances, and the rest served as controls. Second, when estimating the impact of waste disposal, treated households are those that expended some money for waste disposal, while control households did not expend any money for the purpose. A total of 58 households are in the treated group according to waste disposal expenditures, and the rest are controls. Remittance in the study includes both remittances received from within or outside of Bangladesh. Remittance receipt equals 1 if a household receives remittances from within or outside of Bangladesh. The waste disposal expenditure refers to the expenditure incurred on household solid waste disposal in BDT. The waste disposal expenditure variable is constructed from section 9 part-C of HIES regarding monthly nonfood expenditures. Household solid waste disposal is generally performed by third-party community-based waste disposal companies. In addition, monthly health expenditures are divided into inpatient and outpatient expenditures. Total outpatient expenditure comprises consultation fees, medicine costs, diagnosis costs and health-related transportation costs. Total inpatient expenditures include operation charges, bed charges, maternity costs, inpatient consultation fees, inpatient medicine costs, inpatient diagnosis costs, informal tips, formal charges and inpatient transportation costs. The expenditures are measured in BDT (1 USD=109.00 BDT as of December 2023).

To represent the pretreatment characteristics of a household, various relevant explanatory variables, including household size, land ownership, average age, number of household members aged less than 15 years, number of household members aged less than 40 years, and gender of the head of household, are used in the present study. These variables are expected to affect remittances, waste disposal expenditures and outpatient expenditures [25]. Table 1 defines all of the abovementioned explanatory variables.

Table 1 List of matching variables
Full size table

Identification strategies

Estimation of the causal impact of remittances on total monthly waste disposal expenditures and the impact of better waste disposal on total monthly outpatient expenditures is not straight-forward because remittances and waste disposal decisions are not randomly assigned. As a result, remittances and waste disposal decisions entail self-selection bias, referring to a situation in which treated units self-select to receive treatment. Self-selection bias causes the results to be biased. PSM is used to estimate the average treatment effect on the treated (ATET). PSM aims to imitate randomization in the assignment of treatment by matching units with treatment and without treatment [26]. Thus, PSM can be used to construct the counterfactual outcome by means of matching the treated units to the nearest untreated units depending on pretreatment covariates [27]. To address selection bias, this study estimates the ATET using PSM. PSM is based on the assumption that treated and untreated units have no systematic differences after both pretreatment covariates are matched [28].

The ATET is estimated using PSM following the model:

$$ATET(c)=E\left[{Y}_{1}|T=1,C=c\right]-E[{Y}_{0}|T=1,C=c]$$
(1)

C denotes the set of pretreatment covariates. \(E\left[{Y}_{1}|T=1,C=c\right]\) denotes the expected outcomes of the treated units, while \(E[{Y}_{0}|T=1,C=c]\) denotes the expected outcome of the best matched units. T is a treatment dummy for remittance receipt or waste disposal decisions. The remittance receipt equals 1 if a household receives remittances, while the waste disposal decision equals 1 if a household expends any money for the purpose of waste disposal.

The ATET under propensity score \({\text{P}}(c)\) is estimated using the following model:

$$ATET = \left. {E\left[ {Y_{1} } \right]T = 1,P\left( c \right)} \right] - E[\left. {Y_{0} } \right|T = \left. {1,P\left( c \right)} \right]$$
(2)

PSM is used to remove self-selection bias by comparing, on average, similar observations from the treated and control groups.

Results and discussion

Summary statistics

The summary statistics of the pertinent variables of this study are shown in Table 2. The slum households are divided based on remittance receipt. The table describes the statistics of the two groups. Among all the households, 10.04% were treated by remittances, while 89.96% of the households did not receive any remittances. The treated households received 16511.59 BDT from within the country as a local remittance and 19910.12 BDT from outside the country as an international remittance. The control households received 942.96 BDT on average from Cash Transfer, and the treated households received 812.94 BDT on average from Cash Transfer. The households with remittances spent 3.18 BDT more than households without monthly remittances on waste disposal. In other words, households with remittances spent approximately 5 times more than those without remittance. Moreover, the households with remittances spent 133.37 BDT less on monthly outpatient expenditures than did the households without remittances. In addition, households with remittances spent 26.36 BDT more on monthly inpatient expenditures than households without remittances. Household size, cultivable land in acres, homestead land in acres, household members aged less than 15 years and household members aged less than 40 years were not significantly different between remittance-receiving and remittance non-receiving households. However, the average age of households with remittance is 2.14 years younger than that of households without remittance. Finally, more households receiving remittances are male-headed than households not receiving remittances. Similarities in household size, cultivable land in acres, homestead land in acres, household members aged less than 15 years and household members aged less than 40 years of remittance receiving and remittance non receiving households can be interpreted as a balance between the treated and untreated households, while dissimilarities in the average age and gender of households necessitate the use of causal inference mechanisms such as PSM to address the confounder problem.

Table 2 Descriptive statistics
Full size table

Main results

This section presents the main results of the study. Table 3 reports the ATET of remittance receipts, while Table 4 reports the ATET of better waste disposal. The results show that remittances have a positive and statistically significant impact on total monthly waste disposal expenditures at the 1% significance level, while better waste disposal has a negative and statistically significant impact on total outpatient expenditures. The results include caliper matching, kernel matching and nearest neighbor matching. In Table 3, caliper matching shows that the households that received remittances spent 32.74% more on waste disposal than households that did not receive remittances. Similarly, kernel and nearest neighbor matching show that households with remittances spent 32.66% and 28.77% more, respectively, for waste disposal purposes than households without remittances.

Table 3 Impact of remittances on waste disposal expenditures in BDT
Full size table
Table 4 Impact of better waste disposal on health expenditures in BDT
Full size table

In Table 4, caliper matching shows that households with better waste disposal costs 526.26 BDT (1 USD=109 BDT) less for total outpatient treatment purposes than households without waste disposal expenditures. Similarly, kernel and nearest neighbor matching revealed that households with better waste disposal costs 510.64 BDT and 978.67 BDT less for total outpatient treatment purposes, respectively, than households without waste disposal expenditures. In contrast, all the results suggest that better waste disposal does not have a significant impact on total inpatient treatment.

Balancing test

To check the robustness of the results, a balance check between the covariates of the treated and control groups in terms of remittance receipt is conducted. Table 5 shows the balancing property for the treated and control groups before and after matching. It shows that the means of all the covariates except land ownership are significantly different between the control and treated groups before matching. However, after matching, the means of the covariates are not significantly different.

Table 5 Balancing property for the treated and control groups in terms of remittance receipt
Full size table

In addition, Fig. 1 depicts propensity score graphs of the covariates. The graph shows that the distribution of covariates between the control and treatment groups is not homogeneous before matching, but after matching, the distribution becomes homogeneous.

Fig. 1
figure 1

Distributions of covariates before and after matching in terms of remittance receipt

Full size image

Robustness check

To check the consistency of the findings, inverse probability weighted regression (IPWR) adjustment estimation is applied. IPWR is considered a solution if there is a misspecification in PSM [29]. The IPWR result shows that remittance receipt increases waste disposal expenditure by 32.88% at 0.01 significance level.

The IPWR results presented in Table 6 are also consistent with our main results. The results show that better waste disposal reduced monthly outpatient expenditures by 547.70 BDT at the 0.01 significance level, but the impact on monthly inpatient expenditures was not conclusive.

Table 6 Impact of better waste disposal on health expenditures in BDT
Full size table

We conducted ordinary least squares estimation to examine the factors associated with health expenditures. Table 7 reports the results of the ordinary least square estimation conducted with relevant covariates. The results show that family size, total cultivable land ownership, home area land, average age, number of household members aged less than 15 years, number of household members aged less than 40 years, and gender of the head of household are included as covariates. The results show that only the average age and gender of the household head have a significant positive association with monthly outpatient spending. In other words, the number of outpatients spending increases as the average age of a household increases. Furthermore, the results show that only the gender of the household head has a negative association with monthly inpatient spending.

Table 7 Factors related to Health Expenditure in BDT
Full size table

Further, we estimated the impact of cash transfer and micro finance on waste disposal expenditure as alternative financial mechanism. The coefficients of the impact of cash transfer and microcredit receipt are −0.05 BDT and −0.06 BDT respectively. The result shows that cash transfer and microcredit receipt does not increase waste disposal expenditure, suggesting that cash transfer and microcredit are not effective to improve waste disposal expenditure at micro level.

Discussion

In this study, it is found that remittance receipt increases available funds among slum dwellers and, ultimately, increases their expenditures on waste disposal practices. Remittances supplement the income of slum households. They also influence resources by increasing welfare and reducing poverty through macroeconomic and multiplier effects. The willingness to pay for improved solid waste disposal services among urban slum dwellers is dependent on different factors, such as access to solid waste disposal services, the quantity of waste generated per week, the education level of the household, and monthly aggregate income [6]. The investment of households in upgrading environmentally friendly actions such as waste disposal and hygienic behavior is hindered by a lack of access to financing [30]. As such, remittances should be considered a financial mechanism to boost waste disposal activities. Furthermore, our analysis shows that cash transfer and microcredit are not effective at increasing expenditures on waste management.

The increase in waste disposal expenditure can be considered an initiative to improve the environmental conditions of slum surroundings. A small increase in waste disposal expenditure has an enormous impact on improving the health conditions of slum households or decreasing diseases related to unhygienic living conditions. Decreases in diseases related to unhygienic living conditions, such as malaria, respiratory infection and diarrhea, reduce outpatient expenditure [31]. Our finding is supported by the study of Vinti et al. [23], who argued that inadequate waste disposal deteriorates health conditions. This study shows that better waste management decreases outpatient expenditure only. The inpatient expenditure remains unchanged even if a slum household adopts a better waste disposal strategy. The reasons for this can be diverse. First, slum households with low-quality waste disposal often suffer from diseases such as malaria, respiratory infection and diarrhea that do not require admission to the hospital. Outpatient expenditure relates to diseases for which treatment does not require admission to a hospital, and inpatient expenditure relates to diseases for which treatment requires admission to a hospital. Therefore, it is relevant that better waste disposal reduces outpatient expenditures. Second, there can be a spillover effect of low-quality waste management as slum households live in congested areas. Therefore, it can be argued that better waste disposal has only a short-term health impact, as it reduces only outpatient-related diseases and fails to reduce long-term diseases that require inpatient treatment. Furthermore, this study revealed that the average age and gender of the household head are positively associated with outpatient expenditures and that the gender of the household head is negatively associated with inpatient expenditures. In other words, male households have more outpatient expenditures than female-headed households. In contrast, male households have lower inpatient expenditures than female headed households. This finding suggests that health expenditure is associated with the gender of the household head. Moreover, the average age of the household members has a significant positive impact on outpatient expenditures.

This study revealed that remittances are more effective than micro credit and cash transfers in improving household-level waste disposal for several reasons. First, remittances are received without any restrictions, and thus, households can spend money according to their needs, such as waste disposal. In contrast, micro credit and cash transfers are often provided with conditions such as school enrollment, medical checkups, and business. Therefore, microcredit and cash transfers are not effective at enhancing waste disposal spending. Second, the amount of remittances is often greater than that of microcredit and cash transfers. Receiving microcredit involves high interest rates that often negatively affect the socio-economic condition of a vulnerable household, and the amount of cash transfer allowances are very low, which can cause a significant change in the socio-economic condition of vulnerable households. Third, remittance is a bottom-up approach involving the efforts of a household member to improve the life standard of the household and thus is more effective in changing waste disposal behavior. In contrast, microcredit and cash transfers are top-down approaches that often fail to engage household members to improve their waste disposal behavior.

The financial resources that are employed to provide resources to slum households include conditional cash transfer, unconditional cash transfer and micro finance. Cash transfer, both conditional and unconditional, is very limited in coverage, resulting in major changes in the life standards of the whole slum community. On the other hand, micro finance overburdens vulnerable households with the obligation to pay an extremely high interest rate. Thus, slum households can be upgraded not only through government interventions such as cash transfer or financial inclusion but also through self-improvement steps such as remittance earning and income generation [32]. The findings of the study elucidate the approach of making slums self-sufficient through earning remittances to improve waste disposal and hygienic activities. Improving the physical environment of slums by improving waste disposal results in improvements in the health of slum dwellers [33]. Figure 2 shows the findings.

Fig. 2
figure 2

The causal path from remittances to health conditions

Full size image

The effective management of the household solid waste sector holds significant importance since it constitutes the primary source of waste requiring collection within residential communities [16]. The findings of the study have policy implications for develo** countries that struggle to arrange waste disposal financing because of budget shortages for local governments. This study is supported by Afroz et al. [11], who estimated that if each household pays around 13 BDT monthly for waste disposal services, it can arrange 7.6 million BDT monthly in Dhaka city. The waste disposal expenditure incurred by households can be utilized to meet the shortage of funds to finance the high costs associated with improving solid waste management infrastructure [12].

Conclusion

This study suggests that remittances are a financial mechanism for improving waste disposal at the micro level to improve health status. Slum households lack the resources to expend on expenditures such as waste disposal, which can increase their quality of life. This study suggests that bottom-up approaches need to be undertaken to finance household-level waste disposal expenditures that can improve the health status of slum dwellers. Remittance acts in a bottom-up approach, as it provides more sustainable income earned by a family member to a vulnerable household to spend for the betterment of their life. Furthermore, the study suggested that better waste disposal results in decreased outpatient expenditures, an indication of improved household-level health conditions. This study provides practical implications for urban planners on financial mechanisms to manage the waste of slum households and offers evidence that waste disposal reduces outpatient expenditures. The findings of the study can be connected to Sustainable Development Goal 11, which targets sustainable cities and communities.

This study has certain limitations that need to be considered before interpreting the results. First, the study considers waste disposal expenditure as an indicator of better waste disposal instead of the actual amount and quality of waste disposal because of data unavailability. Second, outpatient expenditure is considered an indicator of better health rather than an indicator of actual health, such as the incidence of diseases, because of data unavailability. Third, PSM requires pretreatment covariates as matching variables, as the dataset of this paper is cross-sectional, and it was not possible to find pretreatment covariates. However, this study considers covariates that take years to change, such as household size, land ownership, average age, number of household members aged less than 15 years, number of household members aged less than 40 years, and gender of the head of household. Therefore, these covariates should capture the pretreatment characteristics of the households.