Factors associated with continued Jogini practice in Telangana, South India

Alison Youdlea, Beryl A D’Souza Valib, Nathan Johnc, Pam Andersond

a BMBS, MRCGP, Clinics Director, Good Shepherd Health, India (pseudonym used for security purposes)
b Medical Director, Good Shepherd Health, Good Shepherd Initiative, India.
c MBBS, PhD, Professor of Global Health (pseudonym used for security purposes), India
d Australia.


The Jogini, or Devadasi, system sees young girls “married” to a deity after which she is seen as the property of the village and required to perform religious duties and often sexual favours, typically without payment or freedom of choice. There is a paucity of published research on the factors which make women vulnerable to this exploitation and the factors which increase the likelihood that they are able to extricate themselves from it.

This is a population study of 657 women who had previously been dedicated as Joginis living in 10 villages in Mahabubnagar district of Telangana, South India. The primary outcome of interest was whether the women were practicing as a Jogini at the time of the survey.

Data was analyzed using a mixed effect logistic regression test to determine possible determinants of practicing status. Four factors were found to be significantly associated with continued Jogini practice: 1) being a person with a disability, 2) reason for dedication given as family tradition of other Joginis in the family, 3) poverty in the family of origin, and 4) living in a village with more than ten percent of the population belonging to a scheduled tribe. One factor significantly negatively associated with continued practice was having one or more male children.

Analysis of the demographic data for these women confirmed the previously known fact that the exploitation in the form of the Jogini system disproportionately affects those who are already vulnerable in society — those from scheduled castes (SC) and tribes (ST), other backward castes (OBC), the disabled, the uneducated, and the poor. Between them, SCs, STs, and OBCs make up the non-forward castes, i.e., those who are socially disadvantaged. The scheduled castes, formally known as untouchables, are the lowest of the Indian castes. Scheduled tribes are tribal people, not part of any organized religion. Other backward castes are those between the forward castes and the scheduled castes in terms of social order.

Key words: jogini, vevadasi, sexual exploitation, religious exploitation, disability, poverty, mathamma, India


A Jogini, also known as Devadasi, Mathamma, or various other names across India is a girl or woman committed to theogamy. She forsakes human marriage, usually not of her own volition, and instead is married to a deity. While once widespread across India, the practice of Jogini today is mostly concentrated in three states in a belt across the centre of India: Telangana (formerly part of Andhra Pradesh), Karnataka, and Maharashtra.1 In Mahabubnagar, Telangana, where this study was conducted, the deity is generally the goddess Yellama. This is also the case in various other areas of India.20 The dedication most often occurs just before puberty though it can be much younger.

After dedication, a Jogini’s life is given over to religious duties associated with the worship of the goddess, and this may also include granting sexual favours to any man who asks for it, usually with no recompense. Joginis, today, come almost exclusively from low caste (scheduled caste and scheduled tribe, previously known as Dalit) backgrounds and suffer triple stigma in society – as women, as people of low caste, and as Joginis.2,20 They are ostracized by mainstream society and suffer many disadvantages. Ninety-three percent of Jogini are illiterate, ninety-two percent suffer from depression, and fifty-seven percent have attempted suicide.3 They are ten times more likely than any other Indian woman to die from HIV.4 Nearly seventy-five percent suffer from symptoms suggestive of sexually transmitted infection.2 The majority of Joginis rely on agricultural labour for their livelihood with most earning between Rs 500-1000 (7-14 USD) a month in 2000.3

The tradition of Jogini dates back to the third century AD during the period of the Puranas. There are multiple histories and explanations of the prevalent system. However, it is clear that Joginis were initially of high social status and considered custodians of the arts, especially, music and dance in the temples.4 Their role was to sweep the temple, carry the utensils for worship and dance for the gods as well as perform sex work. However, over time the roles of high and low caste Joginis divided with sex work reserved for the low caste women. Eventually, the practice of Jogini among high caste women died out.3

The justification for sex work is in the ancient Hindu story of Renuka and her son Parasuman. Renuka, an exceptionally pure woman and consort of the sage Jamadagni was beheaded by her son Parasuman at the sage’s behest for lusting after another man. To make retribution for his anger, Jamadagni cut off the head of a passing goddess and attached it to Renuka’s shoulders, at which point she was resurrected. He blessed her and her son, decreeing that from then on, unmarried girls would worship her as their goddess and be ready to satisfy the sexual demands of any man (since Parasuman is present in every man) without expecting anything in return.4 It has been suggested that in more recent times, the religious element can be less important and the practice used to ensure a plentiful supply of women for sex work.20

There have been various movements to outlaw the Jogini system since the 1880s.5 The Indian Penal Code Amendment Act of 1924 (IPC), section 372 prohibits selling minors for purpose of prostitution; sections 372 and 373 declare the practice of dedicating girls for the ultimate purpose of engaging them in prostitution as illegal.20 The first Jogini specific legislation was The Bombay Devadasi Protection Act of 19346 followed by the Madras Devadasis (Prevention of Dedication) Act in 1947,7 effective only in the state of Madras (Tamil Nadu). This was followed by the nationwide prohibition of the practice in 1988, followed the same year by statewide legislation including the Andhra Pradesh Devadasis (Prevention of Dedication) Act.8 Despite this, there have been no prosecutions under the act in Andhra Pradesh. Laws were enforced by the State government of Andhra Pradesh in 2015, giving guidance to officials and power to judges to conduct trials under the act.9

There have been several cases where judges have ruled against the Devadasi system, most notably in Mumbai. Of particular note, the Indian Supreme Court in Vishal Jeet vs Union of India and Ors noted limited progress in implementing laws and called for evaluation of the existing measures by both central and local governments.1 Another ruling by the Supreme Court of India in February 2014 directed the Karnataka Chief Secretary to “take all steps to prevent women being forced to become devadasis at a temple function” at the Uttarang MalaDurga Temple.1 In addition to being illegal within India, Shingal notes nine different international conventions and laws that the practice contravenes, including the Universal Declaration of Human Rights and the Convention of the Rights of a Child.1

The total number of Jogini is unknown; figures obtained from the Andhra Pradesh (AP) government in 2007 estimated 17,000 in the state.10 The AP Social Welfare Commission in 2012 reported 24,273 Joginis in Andhra Pradesh. A subsequent report commissioned by Andhra Pradesh state government and published in 2015 estimated that the true number of Joginis in Andhra Pradesh and Telangana combined could be around 80,000, with another 70,000 in Maharastra.11 Even these figures may be too conservative.

Other than their low caste status, there are many reasons a girl can be dedicated as a Jogini. Suggested reasons in the literature include: physical disability including blindness, deafness, being “crippled,” having no boys in the family (to light the funeral pyre to ensure the father’s entry into the afterlife, or to provide economic support for parents in old age), family tradition (mother or other relative was a Jogini), poverty (an inability to provide a dowry for marriage), or sickness in the family. Religious and superstitious belief is another factor, including pleasing the gods, petitioning for family wellbeing, thanksgiving for the birth of a child after infertility, ensuring fertility of the land, or in place of a human sacrifice.1 The most common factor seems to be if there are already Joginis in the family. One study in Karnataka suggests this was the reason for dedication in 31% of cases followed by a crisis in the family in 16% of cases.2 Both of these factors have religion at their heart.

The literature demonstrates that relatively little is known about why women choose to remain in or leave the system. The closest is a project focused on listening to Joginis themselves conducted by Maggie Black on behalf of Anti-Slavery International in 2007. This revealed the following as Jogini’s perceived problems in life: no permanent husband for economic support, discrimination, superstition, sexual abuse, lack of education, failure of authorities to implement legislation, and child labour (their children).10 These are not necessarily the same factors that keep a woman in the Jogini system, but there may be a close link.

In this study, we look at the differences between practicing and non-practicing Joginis to understand better why Joginis remain?


Study design

Data were collected by field workers between May and June 2016 from 100 villages in Mahabubagar district (as defined by the 2011 Indian census), Telangana, India. The field workers were local men and women working to empower Joginis in other areas of Mahabubnagar district. As locals, they knew the language, dialects, and customs and were trusted by villagers. The 100 villages selected for the survey were identified as potential areas of moderate/high Jogini activity.

Village selection was done using data from a variety of sources. The aim was to select the villages in proximity to each other and the existing project area with the most Joginis. The following data sources contributed to the choice:

Once villages were provisionally selected, field workers attended local Mahela Sangam meetings (monthly meeting of leaders of all self-help groups in the village with government official in attendance), visited the village sarpanch (lay leader elected by villagers – the effective head of village), the village revenue officer (a village level government employee) and ward member (elected representative of the people) and asked them to confirm the data. There was general agreement that these villages were those with a high prevalence of Joginis.

Finally, every known Jogini from the selected villages was asked to complete a survey.

The survey covered basic demographics and questions relating to Jogini tradition, such as how, where, and when they were dedicated, reasons for dedication, respondent’s health and economic status, husband’s income and type of work, number of children, and social and government welfare support received.


The research was verbally explained to each participant and verbal consent obtained to participate in the study. The Jogini community themselves were included in the survey design, and it was introduced to them by trusted members of the community who gave them the opportunity to ask any questions they had. Data was stored securely in locked filing cabinets and with password locked devices. Data for analysis was de-identified by using reference numbers rather than names to identify individuals. Names were not accessible to researchers. The research was deemed ethical by an ad hoc local ethics committee made up of a social worker, doctors specializing in paediatrics and gynaecology, an ex-Jogini, and local leader, and NGO representatives. Informed consent was obtained.

The consensus of the ethics committee was that this data was collected in an ethical manner, and that no harm came to any Jogini as a result of the survey, data collection, or processing.

Data analysis

Our outcome of interest is “practising as a Jogini” at the time of survey. We defined a practising Jogini as a respondent who participated in any of the following activities in a temple or the context of Hindu practice.

  1. holding a basket during a ceremony at a Yellama temple
  2. going into a trance
  3. giving blessings at ceremonies
  4. performing rituals at death
  5. performing at festivals
  6. giving blessings in the name of Yellama

Since women will not admit to sex work, and participation in these activities almost always correlates with those who also still perform sex work, this can be considered a surrogate, although not perfect, marker of ongoing sexual exploitation.

Sample characteristics were analysed using tabulations of factors in the five domains, namely: Jogini’s dedication system, demographic features, socio-economic profile, social support and welfare, and health. Practising status, our variable of interest, was cross-tabulated with factors across the five domains. Possible explanatory factors were either binary or categorical variables, with the exception of age at dedication which was discrete.

To understand the possible determinants of practising status, we conducted mixed-effect logistic regressions, allowing for clustering at mandal and village levels. Each factor in the five domains was rotated into a univariate, mixed-effect, logistic regression test. Based on an initial cut-point of 0.2 significance level, we then selected variables for multivariate analysis which yielded an adjusted odds ratio for each selected variable. These variables, alongside the common confounders, age and income, were included in the final multivariate, mixed-effect, logistic regression test. For mixed-effect, logistic regression tests, age and age at decision were collapsed into binary groups using 35 (age) and 10 (age at decision) as cut-points. Income categories were dichotomised into below or above 2000 IRS (around 24 US dollars) per month. Number of male children was transformed into a binary variable with 2 categories (none or at least 1). Statistical significance level was set at 0.05.

Initially, all data were entered in Excel and subsequently cleaned, processed, and analysed using Stata version 14.


Sample characteristics

The demographic and socio-economic features of all Joginis surveyed are shown in Table 1. Nearly all of the respondents identified themselves as Hindu (99.8 %, n=656), and the majority of respondents came from the scheduled tribes (77.3%). About 80% of them were married, either formally or informally (living long term with one man).

Table 1. Demographic and socio-economic profile of all respondents (n=657)
Demographics Count %
Marital status (n=656) Not married 50 7.62
Married 457 69.66
Married (informally)a 149 22.71
Religion (n=657) Christian 1 0.15
Hindu 656 99.85
Caste (n=656) Backward Caste (BC) 147 22.41
Other (ie., Forward) caste (OC) 1 0.15
Scheduled Caste (SC) 507 77.29
Scheduled Tribe (ST) 1 0.15
Highest level of education (n=655) None 646 98.63
Grades 1 to 7 6 0.92
Intermediate 3 1.37
Socio-economic profile
Monthly income in Indian rupees (n=652) <= 2000 568 87.12
2001 – 3000 76 11.66
3001 – 5000 3 0.46
5001 – 6000 1 0.15
> 6000 4 0.61
Type of work (n=654) Agricultural labour 305 46.63
Manual Labour 326 49.85
Other 23 3.52
Husband’s type of work Agricultural labour 218 45.61
Manual Labour 183 38.28
Other 77 16.11
Husband’s monthly income in Indian rupees (n=507) <= 2000 430 84.81
2001 – 3000 70 13.81
3001 – 5000 3 0.59
5001 - 6000 1 0.20
> 6000 3 0.59
Health (n=656)
Had a physical disability or chronic condition Yes 17 2.59
No 639 97.41
Vision or Hearing Problem Yes 39 5.95
No 617 94.05
Notes. a living with a man; b values are presented as mean and (standard deviation); c n=478

Characteristics of practising Joginis

Based on our definition of outcome of interest (still practising), approximately one quarter (23.5%; 154 out of 656) of the respondents were identified as practising Joginis. The average age at dedication was just under 11 years (10.9 years) in the practising group, slightly younger than that of the non-practising group (11.7 years). However, the age bracket of the practising group was much narrower, ranging between 5 and 25 years, compared to the non-practising counterparts’ 5 and 40 years. All participants were dedicated between 1976 and 2011, with a median year of 2004.

Factors relating to Jogini traditions

Table 2 presents the distribution of Joginis by factors relating to the Jogini traditions. Almost 44% of the still practising Joginis were dedicated because of “all girls in the family” and 35% were dedicated due to poverty. About 18% of practising Joginis indicated that their dedications were either forced by parents (n=15) or induced by superstition (n=12). Nearly 45% (n=68) of practising Joginis were dedicated at a Yellama temple and 37% ( n=57) at other temples. Less than 19% were dedicated at a house (n=29).

Table 2. Distribution of practising status by factors relating to Jogini’s traditions
Still practising (n=154) Not practising (n=502)
Count % Count %
Dedication Reason
All girls in family 67 43.51 285 56.77
Forced by parents 15 9.74 29 5.78
Poverty 54 35.06 127 25.30
Superstition 12 7.79 32 6.37
Other reasons 6 3.90 29 5.78
Other Joginis in the Family?
Yes 17 11.04 22 4.38
No 137 88.96 480 95.62
Place of Dedication
Yellama Temple 68 44.16 112 22.31
Other temple 57 37.01 214 42.63
House 29 18.83 167 33.27
No response 0 0.00 9 1.79

Demographic factors

As shown in Table 3, Hanwada, Jadcherla, and Madhur (mandals/ subdistricts of Mahabubnagar) captured the largest proportions of practising Jogini (40% to 78%); whereas seven out of ten mandals had less than 20% of practicing Joginis.

Table 3. Distribution of practising Joginis by mandals
Mandal Still practising Not practising Total
Bomraspet Count 10 46 56
% 17.86 82.14 100
Dharur Count 8 53 61
% 13.11 86.89 100
Dhaulthabad Count 13 82 95
% 13.68 86.32 100
Gadwala Count 3 54 57
% 5.26 94.74 100
Hanwada Count 26 19 45
% 57.78 42.22 100
Jadcherla Count 39 11 50
% 78 22 100
Kodangal Count 3 63 66
% 4.55 95.45 100
Kosgi Count 3 48 51
% 5.88 94.12 100
Koyalkonda Count 11 68 79
% 13.92 86.08 100
Madhur Count 38 58 96
% 39.58 60.42 100
Column Total 154 502 656

Table 4 shows the distributions of practising Joginis across the major demographic, socio-economic, health, and social support and welfare domains. The vast majority of respondents who were still practising were married, either formally or informally (95%, n=147). Of the 138 practising respondents who were married, 35% had at least 1 male child. Almost all of the practising Joginis had never received any formal education (97%, n=150); whereas the remaining 3% (n=4) had no more than intermediate schooling.

Nearly 20% (n=30) of those who were still practising as a Jogini had a vision or hearing problem compared to less than 2% (n=9) in the non-practising group. Most of the still practising individuals had no physical disability or any self-reported, chronic, health problem (98.7%, n=152). Practising Joginis in our study sample were mostly agricultural labourers (65%), while non-practising counterparts were mainly manual labourers (mostly in construction) or other types of workers. More-89.6%-non-practising Joginis earned less than IRP 2000 a month compared with those still practising as Joginis- 79.2%.

Table 4. Distributions of practising status in various domains
Demographics Practising Joginis (n=154) % Non- practising (n=502) %
Having at least 1 male childa No 90 65.22 241 55.29
Yes 48 34.78 191 44.08
Marital Status Not Married 7 4.55 43 8.57
Legally Married 117 75.97 340 67.73
Married (informally) 30 19.48 119 23.71
Educationb None 150 97.40 495 99
Up to intermediate 4 2.60 5 1
Type of work Agricultural labour 100 64.94 205 41
Manual Labour 47 30.52 279 55.8
Other 7 4.55 16 3.2
Income (monthly)d <= 2000 INR 122 79.22 446 89.56
> 2000 INR 32 20.78 52 10.44
Land ownershipc No 96 62.34 403 80.44
Yes 58 37.66 98 19.56
Social support and welfare
Being a Self-help Group memberb
No 34 22.08 193 38.6
Yes 120 77.92 307 61.4
Having a health cardb No 6 3.90 28 5.6
Yes 148 96.10 472 94.4
Vision or hearing problem No 124 80.52 493 98.21
Yes 30 19.48 9 1.79
Having a physical disability or chronic illnessc No 152 98.70 486 97.01
Yes 2 1.30 15 2.99
Notes. a. n=138 (practising); n=431 (non-practising); b. n=500 for non-practising; c. n=501 for non-practising; d. n=498 for non-practising

Factors associated with practising status

The factors studied affecting whether or not a Jogini continued to practice were divided into 3 domains — economic, social support, and Jogini tradition. Of these 3 associations, several of the factors relating to Jogini tradition were found to have significant association, namely, reason for dedication and other Joginis in the family. No significant associations with practising status were found across the factors in the economic and social support domains.

Table 5. Remarkable differences between still practising as a Jogini and selected demographic, socio-economic health and system-related factors
Unadjusted Adjusted
Explanatory Variable (reference category) Odds Ratio 95% CI p-value n Odds Ratio 95% CI p-value n
Reason for dedication (poverty) 2.19 1.02; 4.68 0.04 656 2.57 1.16; 5.70 0.02 558
Having other Jogini in family (yes) 2.98 1.13; 7.9 0.03 656 2.86 1.0; 8.09 0.05 558
Person with disability (yes) 3.57 1.21; 10.56 0.02 656 6.46 1.80; 23.2 0.004 558
Age at dedication (<=10 yrs) 1.26 0.73; 2.18 0.41 650 1.06 0.59; 1.91 0.84 558
Having at least 1 male child 0.35 0.19; 0.65 0.001 569 0.33 0.18; 0.62 0.001 558
Jogini’s work (agricultural labour) 0.52 0.26; 1.02 0.06 654 0.5 0.25; 1.01 0.05 558
ogini’s age (<=35 yrs) 0.75 0.41; 1.36 0.34 655 0.47 0.24; 0.93 0.03 558
Jogini’s monthly income (l<=2000 IR) 0.74 0.26; 2.12 0.56 652 0.67 0.24; 1.85 0.44 558
Percentage of Scheduled tribes in mandal (>10%) 20.12 3.93; 103.11 0.000 656 10.19 3.10; 33.45 <0.001 558
Notes. Results are based on mixed effect logistic regression allowing for clustering at mandal and village levels with significance level at 0.05%.

Table 5 presents the results of univariate and multivariate mixed effect logistic regression tests. Characteristics of those who were still practising in our study sample can be summarised as: 1) dedicated as a Jogini because of poverty; 2) having at least one Jogini in the family; 3) having no male child; 4) having a sensory disability; and 5) residing in a mandal where at least 10% of the population was from a scheduled tribe. These five variables remained statistically significant after being adjusted for covariates such as age, income and other key explanatory factors in the demographic, traditions relating to Joginis, social support and welfare, and health domains. Of note, the Jogini’s work and age were not significantly associated with practising status in the univariate test, i.e., prior to adjustment, but became significantly associated with remaining as a Jogini after adjusting for key covariates.

Practising status was strongly associated with having a sensory disability. The adjusted odds ratio (6.46) indicated that the odds of still practising in those who were living with a sensory disability were more than 6 times of those without. Respondents who were dedicated because of poverty (OR = 2.57; 95% CI: 1.2 5.7) and having other Jogini(s) in the family (OR 2.86; 95% CI: 1; 2.1) were also more likely to remain active in the Jogini system. The adjusted odds ratio of practising as a Jogini in those residing in mandals with 10% or more of scheduled tribes were 10 times those who were living in other mandals in our sample.

Moderate negative relationships were found between our variable of interest (practising status) and type of work as well as age of respondents. The odds of still practising as a Jogini in those aged 35 years or younger was 50% less than that in those aged over 35. Similarly, the odds of remaining active in the Jogini system among agricultural labourers was 53% less than in those working as labourers in construction work. Having at least one male child was also found to be a protective factor, given that the odds of practising in this group was only one-third of those without any male child.


The main aim of this study was to identify differences between those women dedicated as Joginis who continue to practice compared to those who do not continue to practice. The five factors that were statistically significantly different in those who continued to practice were:

This study is larger than any other published research on Devadasi/Jogini women. Sampark12 interviewed 175 women whilst this study has 656 women. While there have been several studies looking at the social and economic factors at play in the Jogini system, this is, to our knowledge, the first to look at factors which affect whether or not they remain active in the system. Importantly, once a girl is dedicated as a Jogini, she remains stigmatised in the eyes of society for life. However, there are important benefits to her and her children if she no longer practises.

The demographics and socio-economic profile of the group of Joginis in the study was largely similar to that in previously published data. For example, our data shows that the vast majority (78%) of Joginis in the study population are from a scheduled caste (SC). 22.4% were from another backward caste (OBC), the next rung up in the caste system, 0.15% from scheduled tribes (ST) and 0.15% from other (forward) castes. This is unsurprising as it is well known that modern day Joginis are almost exclusively Scheduled Caste (previously referred to as Dalit)6,10 usually Madar (Madiga- an SC group) caste.1 The Sampark Study revealed that Joginis were comprised of 83% Scheduled Caste, 15% other backward castes and 5% general (forward castes) in the Mahabubnagar district where this study took place.12 This distribution is not dissimilar to our data, and the contrast with Sampark’s other data from Karnataka where 100% of Joginis from SC backgrounds is notable. This regional difference may explain the variance between our data’s finding that almost all Devadasi (Jogini) are from SC background since most of the published studies were done in Karnataka. This is important as scheduled caste/tribe status is itself associated with inequality and vulnerability to exploitation with health workers refusing to visit homes in 33% of rural villages and members of the SC/ST community refused access to communal wells in 48% of villages according to a 2006 survey quoted by Torri.13 Hence, Joginis are doubly stigmatised- as Joginis and as low caste women.

The low level of education of the Jogini’s is striking. 98.6% of the Joginis in this study reported no schooling compared to the rural state (Andhra Pradesh) average of only 3.1% of girls aged 6-10 yrs not in school in 2005, the earliest date which the Annual State of Education Report for India was published — the median year of dedication of the women in this study was 2004 at age 10.9 yrs so this is an equivalent comparison.19 It is of course, difficult to know if low education is a risk for being a Jogini or if being a Jogini means they are deprived of access to education. This is consistent with Sampark who found 83% of Joginis in Mahabubnagar were illiterate.12 This is compared with the Mahabubnagar district figures from the National Family Health Survey 2015-6 that found a general literacy rate of 47.2% (published 2019).14

The median age at dedication in this study was 10 years, which is at the top end of the 5-10 year-old, age bracket described by Shingal1 and the middle of Satyanarayana’s15 range of 6-13 years. This young age confirms Shingal’s assertation that “many who perpetuate the devadasi system believe that the younger a girl is dedicated, the more the Goddess will bless her and her family.”1 The current mean age of Jogini’s in our study sample was 37 years. The Sampark Study12 described a typical Jogini as aged 20 to 43 years of age.

The range of dates when the women in this study were dedicated as a Jogini was 1976 to 2011 with a median of 2004. This is significant in the light of the fact that it has been illegal to dedicate a girl as a Jogini since 1988 and shows that the practice was continuing at least until 2011 despite legislation. This finding is corroborated by the Sampark study which also found a large number of dedications after 1988.12 The latest year of dedication may be encouraging as it may indicate that there are few new Jogini’s been dedicated. However, it is also possible that those dedicated after 2011 would be minors and due to legal implications may have been hidden.

The majority (87%) of the Jogini women in this study reported a monthly income of less than or equal to INR 2000 (USD 24) a month with less than 2% earning above the poverty line as defined by the World Bank at USD 1.90 per day. The Sampark study found that 79% of Joginis in Mahabubangar earned less than INR 20,000 per year, the equivalent of INR 1667 per month.12 The low income level of the average Jogini likely makes it difficult to break out of the Jogini system should they wish to leave. Nearly 9% of Jogini women reported a physical disability which was congenital mobility problem (usually club foot), blindness, or deafness in more than two thirds of cases. This is significantly higher than the Indian average of 2.1% of the population,16 suggesting this as another driver to dedication. Black10 and Sampart12 concur, concluding that alongside poverty, ill health is the major precipitant to the dedication of vulnerable girls.

47% of the Jogini women in this study are agricultural labourers and nearly 50% labourers in the construction industry. These are both insecure and low paid forms of daily labour increasing the women’s vulnerability to poverty. This is different from the Sampark study which found 74% of Jogini women engaged in agricultural labour; however, the same study did report a more similar 50% of Joginis doing construction work once they migrated from the village to a city.12 There were two sizeable towns (Jadcherla and Narayanpet) in the vicinity of the villages where this study was done which may explain this.

It is generally accepted that a Jogini cannot marry one man in the traditional sense of the word, hence the popular local saying “a Jogini is the servant of god but wife of the whole town.”. There are no studies that examine marital status of Joginis; the closest is the Sampark study which mentions that the most common family (defined as sharing a home) size among Jogini women in the Mahabubnagar area is 2-4 members (53% of respondents) ; however, they do not specify whether this is children, spouse, or other.12 In our study, a surprisingly high, 69.6% of Joginis, were formally married and another 22.7% informally married, i.e., living with one man. It is important to note that marriage, whether informal or formal, does not necessarily preclude sex work. One possible hypothesis for this finding is the slow reduction in stigma associated with the Jogini system in recent years association with work of both the government and NGOs, since many of the other studies quoted are a decade or more old.

The place of Dedication data shows that approximately 27% were dedicated at a Yellama temple (Yellama is a particular goddess whose cult is known to be associated with the system of Jogini in this area) and another 41% at other temples. Only 29% were dedicated at a house. This is surprising and a little discouraging that dedications have been taking place in public places given that since the outlawing of the system in 1988, there have been efforts by the police to clamp down on dedications. As temples are known places and fewer in number than private homes, law enforcement efforts have focussed on preventing dedications happening there. This finding suggests that such efforts have been only partially successful and ongoing focus in the temples is required.

Pinpointing a single reason for the dedication of a girl child into the Jogini system is difficult since in reality, it is usually multifactorial. There is also an implicit difficulty with this question since those questioned were the Joginis themselves, not their parents who would have made the decision. However, the reasons stated by the women interviewed still gives an insight into the system.

The most common reason given for dedication as a Jogini were that all the children in the family were girls. Our finding of 54% of is similar to the Sampark Study12 that found a similar 57%. This was also stated as a major, though smaller, factor by Black, citing 20% of dedications (study done in the same district).10 There could be some overlap with religious endorsement given as a reason by Black10 as a significant reason for dedicating girls as Joginis was a local religious belief that only a son or a Jogini child lighting a father’s funeral pyre and ensure he has safe passage to the afterlife. Other reasons suggested in the literature for their dedication in this context include: lack of need to find a dowry for the girl’s marriage to a man, hope that a pleased goddess will grant the parents a boy for their next child and need for a child to stay at home to support the parents (traditionally the son’s role).4

Poverty was the second most common reason given for dedication, cited by 35% of Joginis who continue to practice and 25% of those who do not. It is known that Joginis tend to come from poor landless families. Shingal1, Black10, Sampark12 and Satyanarayana15 agree that low socio-economic status should be considered a precipitating factor though this is not quantified in the location relevant to this study.

Nearly 7% (n=44) of Joginis in this study state that the reason for their dedication was superstition. The same number cite that they were forced by their parents. This may be closely linked since the primary reason that the parents force their girls to become Joginis is superstition (although both poverty and having only girls in the family can play a part). Combining the two gives 13.5% of women interviewed citing either superstition or parents forcing them as the reason for their dedication. Sampark12 also state superstition as a major factor in dedications. This may correlate to the 16% of women dedicated as an antidote to a family crisis quoted by Black.10

Other Joginis in the Family was also a reason for dedication. Family custom or a history of other Joginis in the family is known as a significant factor in the dedication of many girls as Joginis. In this study 6% of Joginis had other Joginis in the family. Black10 finds an even higher prevalence of family history of the practice at 31% of Devadasi interviewed, while The Joint Women’s Programme, Bangalore (also Karnataka state) quoted by Shingal puts this at 38%.1 Our study along with these others point towards other Joginis in the family as a strong predictor for dedication of girls in the next generation. It is possible that the differences in our study compared to the Black and JWP studies is accounted for by varying customs in different states — Telangana versus Karnataka.

There are specifics relevant to the geographical areas with high levels or persistence of practice: religious factors and crossover with commercial prostitution. Religious Factors are likely the explanation of the high rates of persistence in Hanwada and Madhur. Hanwada area has one village with a very high proportion of Joginis who are still practicing. This village is the location of an extremely important temple to Yellama and the site of one of the main annual festivals to Yellama. For these reasons, both Jogini and more religious or superstitious people within the general population are drawn to the area. This provides an increased demand for the functions which Joginis fulfil along with likely a better income potential from Jogini work. Madhur is close to both the Karnataka and Maharastra state borders (where the Jogini system is common), has good roads from both and is close to a Yellama temple where there is a lot of Jogini activity and also a large annual festival to Yellama where Joginis play a key role similar to the village in Hanwada described above. Many Joginis migrate to the Madhur area from the adjacent states where the system is also prevalent, and they are less likely to leave as they have “good business” in the vicinity of the temple.

Jadcherla is the largest town in the area and is the “gateway to the city” i.e., the end of the main highway to Hyderabad, 3 hours away, the closest metropolis. All the villages here are essentially part of the urban sprawl of the town. The financial rewards of Jogini work are higher here which would explain why more Joginis stay in the system. There is probably an overlap of commercial prostitution and Jogini practice here, a crossover that was also noted by the Sampark study. They comment that “social stigmatization and better recognition and respect” have encouraged commercial sex workers in some cities to identify themselves as Devadasi.12

Only 23.5% of women surveyed in this study still participated in some practices associated typically with Jogini practice in this area. This statistic is encouraging since it demonstrates that despite the known difficulties many women are able to achieve some degree of emancipation. There have been reports of Joginis who tried to leave being subjected to violence and wilful damage of their property. Despite this, there is ongoing stigmatisation and difficulties for women who have been dedicated as Joginis face irrespective of whether or not they continue to practice.10,11,12 In such close communities, the whole village knows who has been dedicated as a Jogini, and they continue to be considered as second class citizens, and they and their children are discriminated against and remain vulnerable.

There were marked geographical differences in the proportion of Joginis still practicing with prevalence of continued practice ranging from 5% to 78%. Analysis shows a statistically significant association between living in a mandal where more than 10% of the population are scheduled caste, with Jogini women living in these areas having a more than ten-fold chance (OR 10.19) of continuing active Jogini practices. This solidifies the link between caste and not only Jogini practice but the persistence of it.

Women with a disability (vison or hearing problem or had a congenital mobility problem) were at least three and a half times more likely to remain practicing as able-bodied women. A possible explanation is that there is a strong superstitious belief in the area that children with disabilities are in some way marked by the gods for dedication as a Jogini; hence, there is strong societal pressure not only for them to be dedicated as Jogini but to continue to practice. Of course, stigma relating to disability and the fact that they may also find it more difficult to be highly economically active in other ways may also contribute.

Those dedicated because of poverty were more than two and a half times as likely to continue to practice than those dedicated for other reasons. Interestingly, there was no significant association demonstrated in this study between the Jogini’s current income and whether or not she continues to practice. This may be explained by the fact that Indian culture in general and village and Jogini culture specifically are collective cultures where resources are shared with the whole extended family. Therefore, the economic situation of the family of origin continues to impact women throughout their life and limit their opportunities to leave Jogini practice behind them.

When a Jogini was dedicated because she was the only girl child in family, this was negatively associated with continuing to practice as a Jogini. Respondents who were dedicated because of “all girls in the family”, ie., having only girl child in the family, were less likely to continue Jogini practice. A possible explanation for this could be that if the purpose was to allow the girl to light the father’s funeral pyre, there is no need for her to continue as a Jogini after his death.

Just as having another Jogini in the family was a reason for dedication, it was also associated with women being three times more likely to continue to practice as a Jogini than those who did not (OR 2.98 p=0.03). This may be because there is stronger societal and family pressure or acceptance of the practice. Perhaps formative early life experiences may have normalised the Jogini practice is a normal way of life. Joginis tend to live in one room makeshift houses, and much of the sex work involved happens in their home. This means that Jogini’s daughters see their mother’s work from an early age.

In the initial analysis, it was found that having children increased the chances of a woman ceasing Jogini practice. After further breakdown, it became apparent that this effect was only statistically significant if that child was a boy. Having a child has been shown to have a positive impact on a woman’s motivation and ability to stop harmful practises for the good of her child, for example ceasing smoking17 or leaving an abusive relationship.18 The fact that the effect is significantly stronger if the child is a boy may reflect that a male child may improve the social and economic status of the family.

Still going to the Yellama temple was strongly associated with continuing to practice as a Jogini (OR 3.1 p=0.001). This is unsurprising since it is presumably a surrogate marker for those who are still actively involved in the cult of Yellama to whom they were dedicated, a strong justification and driver for the system.

In term of recommendations for eradicating the practice of Jogini and uplifting those families caught up in it, we need to consider the three factors which are both strong reasons for dedication and strongly associated with remaining in the system are:

Women who fit into any of these groups should be considered doubly vulnerable, not only are they more likely to be dedicated, but they are also likely to find it more difficult to leave. These factors can be identified before a girl is dedicated and therefore prevention efforts should be focussed on families who meet these criteria. This should include intensive education on both the harms of the Jogini system and the legal protection available.

Poverty reduction programs specifically focussed on women with a family history of Jogini practice or a disability should be implemented. This should be done on the principles of good community development and could include promoting access to government benefits and entitlements, vocational training for adults, access to loans and business opportunities and schooling for children to reduce their vulnerability to poverty later in life. Intensive and sustained education, ideally by peer educators with credibility in the community should be provided for whole extended families where Jogini practice has been the tradition stressing the harms of the system for future generations and skills to resist societal pressure to dedicate the next generation as Joginis. There should also be determined efforts in both education and poverty reducing measures focussing on areas known to have high scheduled caste populations.

The differences between practising and non-practising Joginis should be interpreted with caution due to the difficulty in ascertaining the practising status from self-report. However, given that the classification of practising status was based on the most obvious signs of practising Jogini, the study results likely reflect some of the common reasons for staying in the system. Second, the study results are not generalisable to other communities where Joginis are present given the sampling method and the regional variations in this practice. The relatively large sample size and the coverage of 100 villages in this study, nonetheless, may provide comprehensive information about the socio-economic profile of Joginis and factors perpetuating the Jogini system in Telangana in India.


This study is the largest study to look at the demographics of Jogini women and the only one to research possible reasons that women once dedicated as Joginis are able to and choose to cease practising. Its findings confirm much of what was already known about the demographics of Jogini women, namely that they are some of the most vulnerable in society even without the sex work which is imposed upon them. It is the first study to look at the place of dedication, finding that more than two-thirds were dedicated in a temple and nearly one-third at home. These findings have significant implications for law enforcement and prevention of future dedications.

Going forward, parallel qualitative studies are needed to understand the changing prevalence and types of Jogini practice in India and the effects of ongoing efforts to eradicate the practice and empower those already dedicated as Joginis. In particular, efforts should be concentrated on understanding where and why new dedications are happening today.


  1. Shingal A. The Devadasi system: temple prostitution in India [Internet]. UCLA Women’s Law J. 2015 [cited 2020 Mar 3];22(1):107. Available from: https://doi.org/10.5070/l3221026367
  2. Dalavi M. Devadasi women - an exploratory study [dissertation]. University of Agricultural Sciences, Dharwad, Karnataka (India). 2010.
  3. Borick J. HIV in India: the Jogini culture. BMJ Case Reports. 2014 Jul 11;2014: bcr2014204635. https://doi.org/10.1136/bcr-2014-204635
  4. Chawla A. Devadasis-sinners or sinned against: an attempt to look at the myth and reality of history and present status of devadasis [Internet]. Samarthbharat. [cited 2022 Sept 23]. Available from: Devadasis - Sinners or Sinned Against (samarthbharat.com)
  5. Kalaivani R. Devadasi system in India and its legal initiatives - an analysis. J Human Soc Sci. 2015;20(2):50-5.
  6. The Bombay Devadasis Protection Act of 1934. No. 10, 171 [cited 2020 Mar 3]. Available from: http://www.lawsofindia.org/pdf/maharashtra/1934/1934MH10.pdf
  7. The Tamil Nadu Devadasis (Prevention of Dedication) Act of 1947. No. 31, 1009-12 [cited 2020 Mar 3]. Available from: http://www.lawsofindia.org/pdf/tamil_nadu/1947/1947TN31.pdf
  8. The Andhra Pradesh Devadasis (Prohibition of Dedicated) Act of 1988. No. 10, 81-85 [Cited 2020 Mar 3]. Available from: http://www.lawsofindia.org/pdf/andhra_pradesh/1988/1988AP10.pdf
  9. AP frames rules to ban devadasi system. Times India. 2015 Dec 10 [cited 2020 Mar 3]. Available from: https://timesofindia.indiatimes.com/city/hyderabad/AP-frames-rules-to-ban-Devadasi-system/articleshow/50115943.cms
  10. Black M. Women in ritual slavery: devadasi, jogini and mathamma in Karnataka and Andhra Pradesh, Southern India [Internet]. Anti-Slavery International; 2007 [cited 2020 Mar 3]. Available from: http://www.antislavery.org/wp-content/uploads/2017/01/women_in_ritual_slavery2007.pdf
  11. One Man Commission Report, V.Raghunath Rao. 2012. Referenced in: NHRC (National Human Rights Commission of India) notices to the Centre and State Governments of Karnataka, Kerala, Tamil Nadu, Andhra Pradesh, Telangana and Maharashtra over continued menace of Devadasi system despite laws banning it; gives six weeks to respond; 2022 Oct 14 [Cited 2023 Apr 22]. Available from: NHRC notices to the Centre and State Governments of Karnataka, Kerala, Tamil Nadu, Andhra Pradesh, Telangana and Maharashtra over continued menace of Devadasi system despite laws banning it; gives six weeks to respond | National Human Rights Commission India
  12. Sampark. Gender based violence on scheduled caste girls: a rapid assessment of the devadasi practice in India. [Internet]. Sampark; 2015 August [cited 2019 September 22]. Available from: Rapid Assessment: Socially Sanctioned Sex Work In Rural Areas: Devadasi Practice (sampark.org)
  13. Torri MC. Abuse of lower castes in South India: The Institution of Devadasi. J Int Women’s Studies [Internet]. 2009;11(2):31-48 [cited 2020 March 3]. Available from: https://vc.bridgew.edu/jiws/vol11/iss2/3/
  14. Government of India. National Family Health Survey 2015-16 [Internet]. Ministry of Health and Family Welfare; 2017 Dec [cited 2020 March 3]. Available from: https://dhsprogram.com/pubs/pdf/FR339/FR339.pdf
  15. Sathyanarayana TN, Babu GR. Targeted sexual exploitation of children and women in India: policy perspectives on devadasi system. Ann Trop Med PH [Internet]. 2012;5(3):157-62 [Cited 2020 March 3]. Available from: https://doi.org/10.4103/1755-6783.98603
  16. Borooah V. Caste, religion and health outcome in India, 2004-14. Econ Polit Weekly [Internet]. 2018 [cited 2020 Mar 3];53(10):65-73. Available from: https://www.epw.in/journal/2018/10/special-articles/caste-religion-and-health-outcomes-india-2004-14.html
  17. Hall ES, Venkatesh M, Greenberg JM. A population study of first and subsequent pregnancy smoking behaviours in Ohio. J Perinatol [Internet]. 2016 Nov [cited 2020 Mar 3];36(11):948-953. https://doi.org/10.1038/jp.2016.119
  18. Estrellado AF, Loh JM. Factors associated with battered Filipino women’s decision to stay in or leave an abusive relationship. J Interpers Violence [Internet]. 2014 Mar [cited 2020 March 3];29(4):575-92. https://doi.org/10.1177/0886260513505709
  19. Pratham. Annual state of education study 2005. Pratham Resource Centre, Mumbai 2005. Available from: https://img.asercentre.org/docs/Publications/ASER%20Reports/ASER_2005/andhrapradesh.pdf
  20. Deane T. The devadasi system: an exploitation of women and children in the name of god and culture. J Int Women’s Studies: 2022;24(1):8. Available from: https://vc.bridgew.edu/jiws/vol24/iss1/8
Peer Reviewed: Submitted 6 May 2023, revised 21 Sept 2023, accepted 23 Sept 2023, published Oct 2023

Competing Interests: None declared.

Correspondence: Beryl A D’Souza Vali, India beryl@gsoim.org

Acknowledgements: With thanks to Good Shepherd Health, India and particularly the Telangana Rural Development Project team for data collection and their deep understanding of the Jogini community.

Cite this article as: Youdle A, D’Souza Vali BA, John N, Anderson P. Factors associated with continued Jogini practice in Telangana, South India. Oct 2023; 10(2):6-21. https://doi.org/10.15566/cjgh.v10i2.769

© Authors. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are properly cited. To view a copy of the license, visit http://creativecommons.org/licenses/by/4.0/