Socio-Economic Determinants of Individual Muslim Zakat Payment Behavior in Indonesia: Analysis of Indonesia Family Life Survey (IFLS) Dataset

. Zakat is one of the pillars of Islam and holds significant potential in Indonesia. However, the collected amount remains below its potential due to the neglect and perceived insignificance of zakat payments by many Muslims. Therefore, understanding socio-economic behaviors and factors influencing individual decisions to pay zakat is crucial. This research investigates the determinants of socio-economic factors on individual zakat behavior, using data from the 2014 Indonesian Family Life Survey (IFLS) focusing on individuals aged 15 and above. Probit regression methodology is employed to examine individual decisions on zakat payments, utilizing STATA 17 for analysis. The findings indicate that individuals with higher socio-economic status are more likely to pay zakat, as evidenced by variables such as gender, household size, age, home ownership, loans, years of schooling, income, religiosity, and occupation significantly influencing contributions. However, variables like place of residence and marital status do not significantly impact payment decisions. Research implications highlight the need for targeted interventions and educational programs to enhance zakat awareness and compliance, especially among individuals with lower socio-economic status.


INTRODUCTION
Islam mandates certain social obligations for all Muslims globally in accordance with Islamic law, and one of these obligations is zakat.Legally defined by Law No. 23 of 2011 on Zakat Management, zakat constitutes wealth that must be contributed by a Muslim or a business entity and given to those entitled according to Islamic law.Indonesia, with the world's largest Muslim population, has broadened zakat objects, encompassing occupational profession, corporate zakat, property zakat, and others, potentially amassing around 327 trillion IDR per year, as indicated by the Ministry of Religious Affairs in 2022.Despite this potential, the Central role, as individuals living nearby are more likely to choose direct visits to contribute (Afifah et al., 2021).Family dependents impact zakat decisions, including homeownership status and the amount of loans (Maulana, 2020).Those with personal homes often have more stable financial statuses, while individuals with loans prioritize income to repay them.
Building on this background, this research aims to examine socio-economic determinants of individual Muslim zakat payment behavior in Indonesia.The selected variables are based on previous research, aiming to address gaps and shortcomings.A significant departure from prior studies lies in the use of secondary microdata, offering a broader reach compared to studies employing primary data obtained through questionnaires and direct interviews (M.Abdullah & Sapiei, 2018;Ghazali & Ibrahim, 2022;Harmaini et al., 2023;Muthi'ah et al., 2021;Afifah et al., 2021;Dianingtyas, 2011;Aulia, 2019).

Zakat
In the Arabic language, "zakat" translates to "growth" or "purification."In terminological context, zakat is an obligation from Allah for every Muslim meeting the minimum nisaab and haul, mandating the allocation of a specific amount for entitled recipients.Zakat evolves into a social obligation in Islamic law, with the recipient (mustahik) experiencing the blessings of Allah's greatness, fostering mutual assistance and love among human beings for the giver (muzaki) (Abdullah et al., 2015) Zakat enhances economic prosperity by redistributing surplus wealth to the poor (Taheri, 2003).A productive zakat can create a better utilization of this levied fund by making the recipients having an opportunity to develop their potential (Yayuli et al., 2021).Alongside prayer (salat), zakat is a mandatory command from Allah SWT for Muslims, serving as a maaliyah ijtima'iyyah worship crucial for community well-being.It functions as both a vertical act of worship towards Allah (hablumminallah) and a form of horizontal worship (hablumminannas), as mentioned in Surah at-Taubah verse 103 of the Quran.
Several factors influencing zakat payment include religiosity or faith level and personal obligations.Religiosity acts as a catalyst for Muslims to pay zakat, fulfilling one of Islam's pillars.Devout Muslims tend to give zakat once they meet the nisaab and haul, irrespective of their income level (Maulana, 2020).Personal obligations, such as debt and home ownership, also impact zakat payments.Some studies suggest that having debt can deter zakat payment,

Consumer Behavior Theory
Within the microeconomic theory framework, consumer behavior becomes the foundation for analyzing individual decisions in paying zakat, utilizing the indifference curve model.In this perspective, zakat is treated as a commodity, and the muzaki (payer of zakat) formulates a demand for zakat, with their utility or satisfaction increasing accordingly (Nicholson & Snyder, 2012).
Muzaki, positioned as a consumer, is also faced with various choices of goods to consume.In this context, zakat is considered a consumption good because it reduces the budget.Therefore, consumption goods are limited by a budget, referred to in the curve model as a budget line.
With budget constraints, muzaki must choose a combination of several goods to maximize utility.In a curve, this combination is simplified into two types, as shown below: The amount of zakat and other items that a muzaki (payer of zakat) can consume is determined by the income variable.The income variable is included in the budget line.Muzaki faces a limited budget, so the amount of zakat and other items they can consume is also limited.If a muzaki has an income, symbolized by I, and this income is allocated to goods x, which is zakat, and goods y, which are other consumer goods, and the price of zakat is Px, and the price of other goods is Py, then the budget line function is as follows: Pxx + Pyy ≤ I …………………………………………………………………………...……………… (1)  In this condition, a Muslim only consumes one type of goods, so their budget is not allocated for zakat.The condition can also be reversed, where a Muslim allocates their entire budget to pay zakat and does not buy other goods.The second possibility is very rare and is not discussed in this research.This study focuses only on the first possibility.

Types of Research and Data
This study relies on secondary data obtained from the Indonesia Family Life Survey (IFLS) dataset, which is a longitudinal socio-economic household survey conducted based on a household sample representing 83% of Indonesia's population carried out across 13 provinces in Indonesia which was gathered by Strauss et al., (2016).The method employed in this study encompasses both descriptive and quantitative research approaches.The variables collected include socioeconomic determinants and factors that influence an individual's decision to pay zakat.Data was processed using the statistical software STATA 17.The data utilized follows a cross-sectional format, specifically the 2014 IFLS 5 and it was analyzed using a probit regression model.This model was chosen appropriately for this research because it offers the advantage of assessing outcomes across multiple categories within the dependent variable, thereby allowing us to determine the probability levels of individuals paying zakat.

Definition of Variables
The variables utilized in this study encompass socio-economic determinants and other factors influencing individuals' decisions to make zakat payments.In total, there are fourteen variables employed in this research.These fourteen variables are categorized into different types: one dependent variable, which is a zakat dummy variable used to analyze individuals' zakat payment decisions, and the remaining thirteen serve as independent variables.These independent variables include gender, age, age squared, household size, homeownership, place of residence, loans, years of schooling, marital status, income, religiosity, and occupation.

Empirical Model
The model in this research is constructed based on theoretical foundations and previous studies.
To examine individual decisions regarding zakat payments, we employ a probit regression model (Mastromatteo & Russo, 2017).This model is utilized when a study aims to explore individual decisions using dependent variables that are categorical or referred to as latent variables (Gujarti & Porter, 2013).Hence, it suits the objective of this research where the dependent variable is either a person pay or not pay the zakat.Probit models are estimated using Maximum Likelihood Estimation (MLE).The general form of the probit regression model is as follows: * =  0 +     +   where i = 1, 2, 3,…,n The variable   * represents a latent variable that serves as the dependent variable for both x and the error term (  ).Meanwhile,   * is an ordinal variable with values ranging from 0 to 1.The equation is transformed into an equation with a dummy dependent variable.
Because of the symmetric normal distribution, we can write this model and use Maximum Likelihood: In the probit model, the coefficients cannot be directly interpreted; instead, interpretation is based on the marginal effects of each individual variable, as follows: 96

Descriptive Analysis
The sample used in this study consists of individuals aged 15 years and above who adhere to the Islamic faith.By utilizing the dataset provided by Strauss et al., (2016) there were a total of 17,547 observations in this study.Table 2 reveals key demographic and socio-economic characteristics of the respondents.The data highlights a pronounced inclination toward zakat payment, with approximately 97.91% of respondents identified as zakat contributors, while only 2.09% do not participate in zakat contributions.This statistic underscores a prevalent practice of charitable giving within the surveyed population.In terms of household size, the average number of individuals per household is approximately 4.15, with a range from a minimum of 1 to a maximum of 15.
These figures provide insights into the typical household composition within the studied population.Regarding gender distribution, the dataset indicates that around 54.39% of respondents are male, while 45.61% are female.This distribution reflects a relatively balanced representation of both genders in the sample, indicating gender diversity among the respondents.The average age of the respondents is approximately 39.07 years, with ages ranging from a minimum of 15 to a maximum of 93.This age distribution offers a glimpse into the demographic diversity and age range of the study participants.In terms of home ownership, the data reveals that 76.79% of respondents own their homes, while approximately 23.21% do not have home ownership.These statistics illustrate the prevalence of home ownership within the surveyed population.
The variable related to the place of residence indicates that 57.0% of respondents reside in urban areas, while 43.0%live in rural areas.This distribution reflects the urban-rural divide among the study participants.Further insights into the financial status of the respondents are provided through the variable related to loans, which indicates that 41.16% of respondents have loans, while 58.84% do not have any outstanding loans.The variable "years of schooling" demonstrates an average of approximately 8.70 years of education, with a minimum value of 0 (indicating no formal education) and a maximum of 16 years.These statistics shed light on the educational attainment levels within the sample.Marital status is characterized by a nearly equal split, with 49.98% of respondents being married and the remaining 50.02% categorized as unmarried.This balanced distribution reflects the diverse marital statuses of the study participants.The variable "income" indicates an average per capita expenditure of approximately 3.17, ranging from a minimum of 1 to a maximum of 5.This data provides insights into the financial resources available to the respondents.Finally, in terms of religiosity, most respondents, approximately 78.98%, are identified as religious, while the remaining 21.02% are classified as non-religious.This variable shed light on the religious orientation of many individuals in the sample.In the context of occupational status, most respondents fall into the categories of self-employed (34.47%) and laborers (39.42%), while the remainder belong to categories such as unpaid family workers and freelancers.

Probit Regression Result
Based on the probit regression results shown in Table 3, it can be concluded that, overall, the independent variables have an impact on individual decisions to give zakat.This statement is supported by the prob>chi-square value of 0.000, which is smaller than the alpha level of 5% or 0.05.The pseudo R 2 value is 0.086, indicating that the combination of all independent variables can explain the dependent variable by 8.6%, while the remaining 91.4% is explained by other variables not included in the model.Microdata research typically involves a larger number of observations or compared to research using standalone surveys or macro-level data.
A larger number of observations mathematically means expanding the denominator in the R 2 , resulting in a smaller R 2 value.However, this study faces the challenge of the unclassified nature of zakat data.In the future, it is hoped that survey institutions such as RAND and BPS Indonesia can enhance modules like IFLS or other surveys like SUSENAS by adding specialized sections on religious matters, including expenditures related to zakat, infak, sedekah, qurbani, umrah, hajj, and others.This would enable other researchers to address the phenomenon of religious behavior among the Indonesian Muslim population through a scientific approach supported by robust statistical data.Additionally, since most respondents in the IFLS dataset are zakat payers, it might impact the inference.Hence, further research should aim to cover this limitation.

Figure
Figure 1.Indifference Curve Source: Nicholson & Snyder (2012)In Figure1, an indifference curve connecting goods x and y is shown.Good x represents zakat, while good y represents other consumption goods.The indifference curve itself is denoted as U1 and has a negative slope.This negative slope indicates the presence of the Marginal Rate of Substitution (MRS).This means that if a muzaki increases the consumption of good x (zakat),

Figure
Figure 2. Budget Line Source: Nicholson & Snyder (2012)From the curve presented at Figure2, the allocation of a muzaki's budget for giving zakat and buying other goods is influenced by the size of their income.The budget line has a negative slope, which is (Py / Px).Therefore, when a muzaki allocates more of their budget to zakat, they will allocate less budget to other goods, and vice versa.The size of the combination of zakat and other consumer goods is depicted in the gray area below the budget line curve.The size of the combination cannot exceed the curve because of insufficient budget.

Table 1
provides definitions for all these variables.

Table 1 .
Variable Definition Place of Residence Dummy variable for the location of the respondent's residence.Takes the value 1 if urban and 0 if rural.
LoansDummy variable for loans.Takes the value 1 if the respondent has a loan and 0 if not.Years of SchoolingThe duration of education completed by household members.

Table 3 .
Probit Regression & Marginal Effect ResultTable above presents the results of probit regression analysis and its corresponding marginal effects.The regression result showed that the prob>chi-square is 0.000.A prob>chi-square value smaller than alpha 5% (0.05) indicates that the overall independent variables have an impact on the behavior of Muslims in giving zakat.Then, after conducting probit regression and its marginal effects as shown in the table below, it is true that a Muslim place of residence and marital status do not affect their behavior in giving zakat.This statement is based on the These factors include the number of household size, gender, age, homeownership, loan amount, years of schooling level, income level, religiosity, and occupation.The findings suggest that the number of household size, as well as gender, age, income level, religiosity, and occupation, are important factors that influence an individual's behavior in paying zakat.However, the place of residence and marital status do not have a significant influence on zakat payment.These findings carry important implications for zakat institutions in Indonesia.Zakat institutions can use socio-economic research to develop strategies to maximize zakat fund collection.For example, zakat institutions can develop programs to raise awareness among individuals with higher incomes, as they are more likely to have surplus funds and can contribute more to zakat.Additionally, zakat institutions can collaborate with religious institutions to enhance religiosity and understanding of zakat obligations among society members.Finally, zakat institutions can work with businesses and corporations to promote good corporate governance and encourage employee motivation to pay zakat.
giving zakat.This statement is based on the significance value (P>|Z|) of 0.000, which is smaller than the alpha level (α=5%/0.05).Looking at its marginal effect, for every increase of paying zakat.