The Risk Factor of Domestic Violence in India

Abstract

Background:

It is over the last decade that research in this field of domestic violence has led to greater recognition of the issue as public health problem. The paper aims to study the prevalence of physical, psychological, and sexual violence and potential risk factors of the women confronting violence within the home in India.

Materials and Methods:

A multicentric study with analytical cross-sectional design was applied. It covers 18 states in India with 14,507 women respondents. Multistage sampling and probability proportion to size were done.

Results:

The result shows that overall 39 per cent of women were abused. Women who have a lower household income, illiterate, belonging to lower caste, and have a partner who drinks/bets, etc. found to be important risk factors and place women in India at a greater risk of experiencing domestic violence.

Conclusion:

As India has already passed a bill against domestic violence, the present results on robustness of the problem will be useful to sensitize the concerned agencies to strictly implement the law. This may lead to more constructive and sustainable response to domestic violence in India for improvement of women health and wellbeing.

Introduction

The ubiquity of domestic violence (DV) can be gauged from the fact that it has been documented in different cultures and societies all over the world. There is growing awareness that DV is a global phenomenon and is a serious issue in developing countries as well.(1) Nevertheless, DV shows particular forms and patterns depending on the local context and recognized as an important public health problem. Despite the range of abuse, it is the most common cause of nonfatal injury to women, who suffer, blame themselves, and choose not to report it. In fact, often rationalize and internalized the abuse by believing that the act was provoked by the women, therefore, justify and accept it as their fate, to continue living with it.(2) The substantial consequence for women's physical, mental, and reproductive health and ultimately the risk of death from DV is also reported to be high, which is committed by a spouse or partner.(3–5)

The prevalence of DV in India ranges from 6 per cent to 60 per cent,(6) with considerable variation across the states in different settings.(3,7,8) However, the magnitude, extent, and burden of the problem in the country have not been accounted well, as the reporting to the problem is still inadequate. In India, few community-based microlevel studies(4,9) are available, which confine to physical violence but evidence on psychological violence and sexual violence is limited.(4) There is also very limited empirical evidence of its various determinants, outcome, and their relationships.(10)

Various studies from South Asian countries on DV have identified a number of associated individual and household level risk factors which shows that certain demographic factors such as age, number of living male children, and living in extended family have an association with DV.(11,12) Among the protective factors identified in developing countries are higher socioeconomic status, women's economic independence, quality of marital relationship,(9) and higher levels of education among women.(13,14) The risk of spousal violence against women is globally known to be higher among women who are younger, have a lower household income, less educated, belonging to lower caste, nonworking women, partner who drinks/bets, etc.(4,8) However, the issue of DV and its underlying social determinants of DV in developing countries remain limited especially in the context of India.

This paper tries to study the prevalence of physical, psychological, and sexual violence and its potential risk factors for women respondents with their background characteristics such as age, religion, caste, education, occupation, and income, and its association. The term DV is used in the article refers to the violence faced by the women from their husband and family members within the marital home. Any form of DV includes physical, psychological, or sexual violence faced by the women.

Materials and Methods

Study design

It was a multicentre study and the study design was conceived as an analytical cross-sectional study. Both quantitative and qualitative methods were used. A population-based approach was applied to find out the association between DV and reproductive health consequences.

Population

Inclusion criteria were the married women in the age group between 15 and 35 years. Exclusion criteria were unmarried, widow, and separated women.

Sampling frame

The study was carried out in all the six zones of India, i.e., Northern, Southern, Eastern, Western, Central, and North East, to have a wider representation. Based on the prevalence rate from NFHS-2,(2) the states with high, medium, and low prevalence of DV were selected. In total, 18 states were randomly selected. Keeping in view of 70:30 ratio of rural and urban population, the samples were distributed accordingly. Multistage sampling strategy was used to attain the required samples. For rural sample, two districts and two blocks were selected randomly. 124 villages were chosen for women participants randomly. For urban sample, district headquarters were considered and three socioeconomic strata were identified as high-, middle-, and low-income groups. To select the married women from urban and rural areas, a systematic sampling procedure was followed for households.

Sample size

The sample size was calculated based on the available study that the bad obstetric outcome of pregnancy was 8 per cent and it was expected that the risk would be double (OR = 2) with women subjected to abuse or violence. Using WinPepi, a total of 14,405 female samples were considered for the study (Alpha = 0.05 and 1-Beta = 0.80), which included a margin of 10 per cent nonresponse. Probability proportion to size was calculated for each state.

Study instrument

The study involved collecting data through semistructured questionnaire. A multiphase process was used to develop these questionnaires to ensure that it was culturally and linguistically appropriate. The questionnaire was prepared initially in English and translated and back translated to ensure semantic and content validity. The translated questionnaires were further reviewed for linguistic reliability and appropriateness by the field investigator.

Data validation and management

The data entry package (Epi 6) and the tabulation plan were sent to each centre to bring uniformity. After receiving the data from six participating centers, data were merged. The data were cleaned and validated using excels double data entry.

Data analysis

The data analysis were done using Epi Info, transport to SPSS to calculate proportion, OR, and multivariate logistic regression. 95 per cent confidence intervals (CI) and a P value of less than 0.05 were considered as the minimum level of significance. Content analysis was done for the qualitative data like Focus Group Discussion (FGD), in-depth interview, and case study, respectively.

Measurements

The factors associated with DV included for the analysis were individual- and community-level variables. Multivariate analysis using binary logistic regression (forward method) was applied to 14,507 cases. The statistically significant (<0.20) variables observed in the univariate analysis were included for multivariate analysis. For logistic regression, these variables were used as categorical variables, except the age which was taken as continuous variable. The final model got stabilized after undergoing 12 iterations. Overall efficiency of the model was found 87 per cent approximately. The most parsimonious model obtained in the multivariate analysis with 14,507 cases of which 13,951 have been included in the analysis and 556 cases were missing.

Ethical consideration

The study was approved by the Human Research Ethics, ICMR, New Delhi, and reviewed by senior staff for cultural appropriateness. Informed consent was obtained from all participants, participation was entirely voluntary and confidentiality assured.

Result

The study data revealed that DV was very much prevalent irrespective of rural-urban differentials in the country. On the whole, 39 per cent of the women have mentioned about the incidence of one or the other forms of DV in all the six zones. However, overall 37 per cent of the women indicated prevalence of psychological violence, about 14 per cent of physical and sexual violence in their homes, respectively.

Risk factors of DV

The potential risk factors associated with DV reported by the respondents are discussed using multivariate logistic regression analysis [Table 1].

Table 1.

Binary logistic regression: Risk factors of domestic violence

Age

It is evident that women in the age group between 21 to 35 years were significantly at one time risk of facing DV compared to the women belonged to less than 20 years of age. However, there was a slight decline after 30 years and above age group, which was quite expected as women of higher age group were bound to reduce violence with the passage of time by virtue of their position betters with having adult sons in the family.

Religion

The data revealed that women belonged to Muslim religion were at more risk of facing any form of DV compared to women belonged to Hindu religion. While women belonged to Christianity and Buddhism were at no risk, depicting the religion being the protective factor.

Caste

The analysis reveals that the infliction of physical as well as psychological and sexual violence was most prevalent among lower caste women who were significantly at greater risk of facing any form of DV compared to upper caste groups.

Education

The data reveal from the regression model that illiterate women were two times significantly at risk of DV (OR = 2.112, CI = 1.812–2.461), whereas women who have completed up to 10 years of schooling (OR = 1.703, CI = 1.474–1.968) and graduation or higher education (OR = 1.207, CI = 1.052–1.384) were significantly one time at higher risk for injury from DV, respectively, compared to the women having professional degree. Though violence decreases with increase of education, the magnitude of DV was considerably high among women with higher literacy also.

Occupation

The occupation of the participant was recorded and the responses were categorized into i) working women, those contributed to the household income in terms of cash may be engaged in small businesses, daily-waged skilled and unskilled laborers, etc., and ii) house wife. Out of the total women working in different sectors, 49 per cent were facing DV compared to the housewives (36 per cent). In contrast, women who contributed financially none (the house wives) than women whose earnings contributed more to covering their household's expenses were significantly (OR = 0.735, CI = 0.669–0.807) less at risk for DV. Across all the zones, prevalence of DV was higher among the working women compared to the homemakers which were quite contrary to the expected norm. Intraoccupational comparison reveals that women working as nonskilled laborer were facing more DV than the working women of other sector.

Family income

The income of the respondents is indicated by the household's net income per month. The income details were collected in Indian Rupees (INR). The association of family income and DV was found to be highly significant. Women fell in the category of monthly income up to Rs. 3000 were at one time risk of DV compared to women in the family income of Rs. 3001-Rs. 5000 and above.

Size of the family

The association of size of the family and DV was found to be highly significant as women belonged to the family size of the 5–7 members and more than 8 members had one time risk of facing DV, respectively, than the women belonging to the smaller family size of 2 and 3–4 family members, respectively.

Type and length of the marriage

The marriage was categorized in three types namely, arranged marriage, love marriage, and mix marriage, which was love marriage settled by elders. The proportion of the women who reported experiencing DV was significantly two times higher among the women with arranged marriage and three times higher among the women with mixed marriage, respectively, than among women with love marriages. The result shows that any form of DV decreased as the space of marital life increased.

Alcohol consumption

It is evident from the regression model that the prevalence of DV was significantly two times more where husband was found alcoholic (OR = 2.556, CI = 2.358–2.771) as compared to women whose husbands were not habitual of alcohol. However, alcoholism might not be the sole cause of DV as DV was also reported in homes where husband was reported nonalcoholic.

Bet and gamble: Gambling was another menace which leads to DV. It was found from the model that women whose husbands were in the habit of betting and gambling were significantly five times higher at risk of DV (OR = 5.869, CI = 4.561–7.553) as compared to those women whose husbands were not having such habits.

Discussion

In the present study, the prevalence of DV in India was considerably high persisting across all socioeconomic strata existing in all the communities.(5,11) Empirical results have suggested that education of women have an association with DV, which reflects a shift in the thinking pattern and burgeoning down the balance of power between husband and wife.(9,15) But, the odds of DV were reduced only for women who had achieved higher education, suggesting that modest increases in educational attainment available to the majority of the women in India will not substantially alter their risks. However, data reflect that the victims were not only among illiterate and poor, who were besieged in traditional folklores and customs, it occurs across all social categories and social set-up.(5,10) However, results reported that women working and contributing to the household budget were at increased risk of violence. The expectation expressed in the qualitative data that women's participation in economic activity would lead to higher status, security, and as a protective buffer against DV appears less realistic in the light of the quantitative results.(1,3)

One limitation of the study was that the family income was calculated based on self-reported items produced in the agricultural land. We presume that the women may not report correctly due to stigma and embarrassment. Previous studies suggest that highly normative support for violence against women exists in this setting and therefore may lead to underreporting Stephenson et al. 2006.

Despite the limitations of reporting bias, the findings highlight the complex and often contradictory nature of the relationships among factors at different levels and the ways in which they influence women's risk of suffering DV. In this context of gender inequality and poverty, the practice of patriarchy appears to exacerbate women's risk of DV. These causes reflect deep-rooted gender inequalities that persist across India.(9,14,15) The findings of the association between the above analyzed factors suggest that there are broader and overarching reasons behind DV, whose implications go beyond individual and psychological situations. This practice of interpersonal violence may lead to affects the health of the women.(1,11) Recognition of emerging health issues is needed to address women facing DV within the cultural milieu to improve maternal health and well-being.

The appalling toll will not be eased out until family, government, institutions, and civil society organizations address the issue collectively. These results provide vital information to assess the situation to develop interventions as well as policies and programmes toward preventing violence against the women. As India has already passed a bill against DV, the present results on robustness of the problem will be useful to sensitize the concerned agencies to strictly implement the law.