Working the Pony

Abstract

We present a model of the drinking practices of Latina immigrants who work in bars or cantinas and are compensated for drinking 8 oz beer or ponies with customers, hence working the pony. A total of 358 women recruited from 61 randomly selected cantinas were interviewed as part of a community study conducted in a metropolitan area in the U.S. Southwest. Participants reported drinking an average of 21 beers on nights they worked for ponies. Multiple regression results indicate that drinking practices are related to sociodemographic, work history, drug use, and sexual risk characteristics. The drinking of cantineras is more than 4 times higher than the heavy episodic drinking of high risk Latino male drinkers and young male restaurant workers. The findings suggest that they may be at risk of multiple health problems and could benefit from alcohol risk-reduction interventions.

Keywords

alcohol abuse drinking practices Latinas women health bar workers

Drinking alcohol as an occupational activity has rarely been reported in the research literature, even though accounts of this practice are generally known among public health practitioners and reported in the popular media (Hewitt, 2010). The sparse research literature on this topic indicates that cantina-specific drinking practices are common in the United States (Ayala, Carrier, & Magaña, 1996) and specifically in Texas and other parts of the Southwest (Ferreira-Pinto, Ramos, & Shedlin, 1996). Cantineras are a subgroup of immigrant Latinas who suffer from the stress of acculturating to the United States in a manner similar to other vulnerable Latina immigrants whose poverty is amplified by their undocumented status, limited education, and poor command of English (Herrera & Campero, 2002). They also share with other Latina immigrants an initial job history in the United States working in unskilled occupations mostly in the service sector, where they are employed cleaning houses, babysitting, and working in restaurants (Hondagneu-Sotelo & Avila, 1997). Unlike other immigrants, however, they opt for a job in cantinas after experiencing poor treatment and unsafe conditions in a previous job (Fernández-Esquer, 2003). This choice of occupation sets them apart from other Latinas’ immigration experience, as their choice of work in local cantinas (i.e., working class Latino bars) increases their risk of adverse social and health consequences. Their choice of bar work is also influenced by coworkers and friends, and it is based on the premise that it will bring a better salary compared with other jobs available to them. As the study results will later reveal, cantineras do not make real gains in economic power, and their attempts at making a better income in a cantina come at great health, social, and personal costs.

Women working in cantinas come primarily from Mexico and Central America and are called cantineras, which in Spanish means “employee of a cantina” according to local usage of the word. These women earn a living by working as waitresses, many of whom earn sales commissions based on the beer they drink, consumed as part of an entertainment contract, acquired by and paid for a male bar customer. This “drinking for wages” phenomenon has been reported previously (Fernández-Esquer, 2003), as well as the role that stigma plays in the occupational performance of cantineras (Fernández-Esquer & Agoff, 2012). Previous studies conducted among Latina women who work in bars have documented their HIV-related sex risk practices (Ayala et al., 1996; Fernández-Esquer, 2003; Ferreira-Pinto et al., 1996) but their drinking at work has not been explored. In this report, we characterize drinking as an occupational activity and economic survival strategy embedded in the sociocultural Latino context of a highly marginalized occupation prevalent in local working class bars.

Alcohol Use Among Latinas

The drinking of cantineras is atypical. Studies describing drinking patterns in the United States generally characterize Latinas as a population of light and infrequent drinkers (Finch, 2001; Treno, Alaniz, & Gruenwald, 1999). The recently immigrated fit this description particularly well (Caetano, 1994; Canino, 1994; Zemore, 2007). As U.S. Latinas become more acculturated, educated, and affluent, their drinking levels increase (Parker & Weaver, 1995; Zemore, 2005), sometimes surpassing non-Hispanic Whites in rates of past year alcohol use and reported alcohol problems (Farabee, Wallisch, & Maxwell, 1995; Zemore, 2007). Marital status influences drinking behavior, as Latina women who are separated or divorced report higher drinking levels compared with those who are single or married (Treno et al., 1999).

The heavy drinking context in which cantineras work may also place them at greater risk of substance abuse. The drinking of cantineras is more than four times higher than the heavy episodic drinking estimated for high risk Latino male drinkers and young male restaurant workers (Kjaerheim et al., 1995; Moore et al., 2009). Research indicates that even when Latina immigrants have low rates of participation in substance use including cigarette, marijuana, and cocaine, those exposed to high-risk contexts are more likely to engage in these behaviors. A study completed in California found that 17% of recently immigrated Latinas surveyed in high-risk behavior venues such as bars, clubs, and work camps reported using cocaine or methamphetamine (Hernandez et al., 2009). Cantineras exposed to these drugs through their work may be more likely to consume them particularly if drinking heavily.

Social Context

Many immigrant Latinas leave families behind in their countries of origin and find themselves without a supportive social network after their arrival to the United States. The absence of significant others not only leads to loneliness and isolation, but when economic hardships are present, their absence also creates the opportunity to venture beyond traditional occupations and into jobs that violate social network and gender expectations and promote health-compromising behaviors (Fernández-Esquer, Diamond, & Atkinson, 2010; Umberson, Crosnoe, & Reczek, 2011). The drinking of cantineras deviates from Latino social and drinking norms that discourage women from heavy drinking (Zemore, 2005). From a traditional Latino culture perspective, a woman who drinks in a bar patronized only by men is violating three gender norm expectations: (1) Women should not drink heavily, (2) in the company of male strangers, and (3) in a public establishment designated as the social domain of men. Consequences of violating these gender norms may include stigma, discrimination, and the loss of their social standing as “decent women” (Fernández-Esquer & Agoff, 2012). They are also likely to be regarded by the local community as disreputable, and, in a manner similar to other women around the world, they are perceived and treated by men who visit these establishments as easy sexual targets (Fernández-Esquer & Agoff, 2012; Norris, Kitali,&Worby, 2009).

Drinking for Wages

The drinking patterns reported in this article are based on the findings of a large community study with cantineras (Fernández-Esquer & Cardenas-Turranzas, 2004) and on an ancillary qualitative study of their alcohol practices (Fernández-Esquer, 2003).

The term “drinking for wages” describes cantineras’ consumption of alcohol in exchange for payment. Cantineras hired as “hostesses” interact with and get paid for drinking beer with male customers. Hostesses may work for beer commissions (temporary hostess), or for a salary supplemented by beer commissions (see Fernández-Esquer, 2003, for more details). The expression “working the pony” (“buscar la pony” in Spanish) describes the act of looking for the opportunity (i.e., hustling) to make money by drinking ponies sold at a premium to customers of the bar. A “pony” is an 8 oz. beer commonly consumed by cantineras to fulfill job duties that will be described in more detail in the “Results” section of this article.

In this article, we model the drinking patterns of cantineras and explore the extent to which demographic background, cantina work history, substance use, and sexual behavior are associated with their heavy drinking practices. Based on previous research among Latinas, we hypothesize that heavy drinking among cantineras will be more common among the younger and less educated compared with their counterparts. Contrary to the findings of previous studies, we do not expect that drinking will increase with greater length of residence in the United States or length of employment, as working in cantinas is a job option primarily for recently arrived Latinas, and their job demands and economic pressures require adapting quickly to high-volume drinking. As cantineras working as hostesses drink with bar customers to increase their income, their reported “tickets” will be positively related to monthly income. We also predict that the more a cantinera drinks, the more she is likely to participate in sexual risk-taking activities such as having multiple sex partners, having sex after drinking, and trading sex for money.

Method

This study is based on secondary data collected over the course of a community intervention study conducted in 1998 and titled “GIRASOL (Sunflower): Community Outreach to Prevent Cervical Cancer among Latinas” (Fernández-Esquer & Cardenas-Turranzas, 2004). GIRASOL was funded by the National Cancer Institute to design, implement, and evaluate a community intervention program to promote primary and secondary cancer prevention among cantina workers. While the focus of the parent study was cervical cancer, formative data collected among cantineras revealed the importance of alcohol use as a work practice and its potential as a barrier for cancer screening. All data collection and intervention procedures adopted in this study were approved by the Committee for the Protection of Human Subjects at the University of Texas Health Sciences Center at Houston.

Participants

Participants in this study were recently immigrated Latina women from Mexico and Central America who were employed in working class bars located in neighborhoods predominantly populated by Latino residents (<80%) in a large metropolitan area in Texas.

Participant recruitment

Bars located in Latino neighborhoods were observed and enumerated for potential inclusion in this study. Ninety-three cantinas were identified as eligible, and of this total, 61 were randomly selected for inclusion in the study. The remaining 32 cantinas were eliminated due to refusal (n = 4) or because they were no longer needed once the bar quota had been reached (n = 28). After receiving approval from the owner, the project coordinator approached the bar workers and scheduled a mutually agreed date and time to complete their interviews. On the scheduled date, members of the interview team approached each of the employees present at the bar and requested to complete an anonymous interview. A total of 435 women were approached at cantinas. Of this total, 360 completed an interview and 75 declined to be interviewed (survey completion rate 83%).

Interviewer training

Interviewers’ training consisted of a 1-day training session that presented an overview of the study, reviewed the content of the baseline questionnaire, and role-played its proper administration. Training also consisted of an average of 8 hours of interviewing practice, completed before the first field interview assignment. The project coordinator monitored each interviewer’s pilot interview and reviewed questionnaire coding flagged during data entry the previous week. All interviewers were Latina women familiar with the cantinera community and fully fluent in Spanish. Over 95% of the interviews completed by the team were conducted in Spanish.

Interview procedures

Women working in each selected cantina received a full description of the project during the project coordinator’s initial visit. On a date previously arranged with the bar workers, the team of interviewers visited the assigned cantina and asked each waitress for permission to complete an interview. If a woman granted permission, she was asked to step aside and meet with the interviewer at a table or other public location away from the bar’s activity. A verbal informed consent statement was obtained by interviewers prior to the interview. It explained the purpose of the study and the nature of the interview, including its anonymous and voluntary nature. Each participant received a US$10.00 coupon to be redeemed at a local supermarket.

Measures

The GIRASOL baseline survey was conducted in 1998 and included measures of cancer-related knowledge, attitudes, and behaviors previously tested with other Latina populations (McAlister et al., 1992). Questions related to alcohol use were based on measures routinely used in the Behavioral Risk Factors Surveillance System (BRFSS; Centers for Disease Control and Prevention [CDC], 1991) and on an initial formative study completed in 1996 (Fernández-Esquer & Cardenas-Turranzas, 2004). The following description provides more details about the measures utilized in this study presented in Tables 1 and 2.

Table 1.

Sociodemographic Characteristics of Cantineras Interviewed in 1998 (n = 360).

Variables	%
Country of origin
Mexico	33.1
Honduras	36.2
El Salvador	19.5
Other Latin American	8.1
The United States	3.1
Spoken language
Only/mostly Spanish	92.5
Both equally	3.4
Only/mostly English	4.2
Read in Spanish
Good or very good	71.4
More or less	22.5
Bad or very bad	3.9
Cannot read	2.2
Write in Spanish
Good or very good	69.2
More or less	22.8
Bad or very bad	5.8
Cannot write	.2
Job type
Waitress	22.8
Temporary hostess	27.6
Salaried hostess	49.6
How often has sex after drinking
Quite often	13.4
Sometimes	49.1
Never	37.5
Sex partners met in a cantina
All or almost all	7.6
Some	33.2
None	59.1
Sex trade recency
Never	88.6
>6 months ago	1.8
<6 months ago	2.3
Last 30 days	5.8

Table 2.

Correlations of Drinking Practices With Sociodemographic, Work History, Drug Use, and Sex Risk.

Variables	M	SD	Total days drank last month n = 356	Most beers per single night last month n = 330	Normal day drinking n = 340	“Tickets” on a good night n = 271
M			15.40	21.47	11.19	27.83
SD			7.92	14.4	7.08	17.45
Respondent age	30.2	7.4	−.06	−.14**	−.15**	−.25***
Years of education	6.9	3.5	.01	−.03	.09	−.08
Monthly income	1,131	674	.10	.20**	.20**	.22**^a
Years living in the United States	4.8	6.3	−.06	−.15**	−.12**	−.15**
Trusted relatives	1.5	1.67	−.11*	−.26**	.021	−.21***
Trusted friends	1.2	1.65	−.04	−.20**	−.02	−.14*
Years working in cantina	.97	1.35	.08	.04	.07	.04
Days/week in cantina	4.84	1.32	.31**	.08	.08	.17**
Cantinas’ work in the last 3 years	1.57	.82	.05	.15**	.01	.16**
Waitress job type	—	—	.18**	.19**	.23**	.54**
Total cigarettes per day	5.62	9.42	.10	.20**	.04	.29***
Marijuana use last year	2.32	20.92	.02	−.01	−.03	.15**
Cocaine use last year	2.70	22.28	.03	.01	.06	.17**
Had sex after drinking	1.77	0.71	.21**	.16**	.26**	.18**
Sex partners last 12 months	1.39	.94	.23**	.23**	.08	.18**
Sex partners in cantina	2.45	0.69	.02	.05	.14***	.11**
Traded sex recently	0.33	0.89	.09	.05	.07	.14**

n = 253.

p = .05. **p = .01. ***p = .001.

Sociodemographic characteristics

This group of variables included the respondent’s income, age, years of schooling, and years of residency in the United States. We included two measures of social support related to number of trusted friends and relatives. We also asked questions about the respondent’s country of origin, spoken language preference, and ability to read and write in Spanish (see Table 1).

Cantina work history

To probe for stability of bar employment, we requested information about the length of time working in the current cantina (coded in years), total number of cantinas worked in the previous 3 years, total days per week the respondent worked in a cantina, and the type of waitressing job she performed (“Do you work for a salary [waitress], “tickets” [temporary hostess] or salary and “tickets” [salaried hostess]?”). As the type of waitressing job encodes risk of alcohol abuse, it was coded as a categorical variable where a higher number represents a higher risk: 1 = waitress only, 2 = hostess working for “tickets” only, 3 = hostess working for salary and “tickets” (Table 1).

Sex risk and drug use

Variables measuring a cantinera’s health risks included frequency of having sex after a night of drinking (1 = never, 2 = sometimes, 3 = quite often), sex partners met in a cantina (1 = none, 2 = some, 3 = all or almost all), respondent’s recency of trading sex for money (1 = never, 2 = more than 6 months ago, 3 = less than 6 months ago, 4 = in the last 30 days), total cigarettes consumed per day, total number of times marijuana and cocaine were used in the last year, and total number of sexual partners in the last year (Table 2).

Drinking practices

Variables related to alcohol consumption were derived from the BRFSS and the formative study. Measures of frequency and quantity of alcohol use based on the BRFSS (CDC, 1991) included (1) the total number of days in which alcohol was consumed in the last month and (2) most beers consumed during a single night last month. Formative study measures included the (3) total number of beers consumed during a “normal” day at work (how many beers do you drink in a normal day at work?), a measure of estimated average daily drinking reflecting habitual drinking, and (4) number of “tickets” (how many tickets do you make in a good night at work?) as a measure of total beers consumed by the cantinera at work in exchange for money.

Data Analysis

The objective of the data analysis was to describe the drinking patterns of cantineras and to estimate the relative influence of sociodemographic characteristics, cantina work history, substance use, and sexual behaviors on cantineras’ drinking patterns. As an initial step, we conducted univariate statistics to estimate the association of each of the mentioned predictors and the drinking variables (Table 2). We then regressed the measures of drinking behavior on the demographic and risk behavior variables to test the simultaneous influence and the total variance accounted for by the set of predictors on each drinking variable (Table 3). As the current study is based on a cross-sectional design, the variables entered in the regression equation estimated within-time variation in drinking dependent variables. Consistent with General Linear Test procedures (Wasserman & Kutner, 1983), we tested the relative ability of the predictor variables to significantly reduce error or unexplained variance in variables representing alcohol consumption among cantineras. Each linear regression added all predictor variables to the equation as a single block using the “Enter” procedure, to test their simultaneous influence on each of the alcohol-dependent variables using SPSS Version 10 for Windows.

Table 3.

Multiple Regressions: Sociodemographic, Work History, Drug Use, and Sex Risk Predictors of Drinking Practices of Cantineras in 1998.

	Total days drank last month n = 356	Most beers per single night last month n = 330	Normal day drinking n = 340	“Tickets” in a good night n = 271^a
Variables	Standardized β	Standardized β	Standardized β	Standardized β
Respondent age	−.02	−.09	−.10	−.18**
Years living in the United States	−.01	−.09	−.08	−.13*
Years of education	.03	−.01	.07	.08
Monthly income	−.04	.09	.04	.13**
Trusted relatives	−.05	−.17**	.03	−.13*
Trusted friends	−.06	−.13*	−.09	−.03
Years working in cantina	.10	.20**	.14*	.17**
Days/week work in cantina	.27*	−.05	.01	−.05
Total cantinas last 3 years	.04	.14**	.07	.22***
Waitress job type	.13*	.12*	.18**	—
Total cigarettes per day	.01	.14**	−.01	.23**
Marijuana use last year	−.13	−.19	−.57***	−.07*
Cocaine use last year	.10	.12	.52***	.20
Had sex after drinking	.15**	.11*	.23***	.10
Sex partners last 12 months	.14*	.14**	−.01	.04
Adjusted R²	.16***	.21***	.16***	.25***

Includes temporary and salaried hostesses only.

Results

Demographic and Sex Risk Characteristics

The cantineras participating in the survey (see Table 1) were primarily from Mexico (33%) and Central America (Honduras 36% and El Salvador 19.5%) and an overwhelming majority was Spanish monolingual (92.5%). Most cantineras were hired as salaried hostesses (49.6%) who received a minimum salary, which was augmented with tickets. Respondents reported having sex after a night drinking at work “sometimes” (49.6%) and meeting “some” (33%) of their sexual partners in a cantina. About 11% reported trading sex for money in the last year.¹ Cantineras were in their 30s and had an average monthly income of US$1,131. Their average educational level was elementary school and their length of residence in the United States was slightly under 5 years. Cantineras appear not to trust many people, as they reported an average of only one relative and/or friend to trust.

Substance Use

Current drug use was relatively low, as only 9% of cantineras reported using either marijuana or cocaine in the last year. However, approximately 41% of cantineras were current smokers, 15% had tried marijuana, and 10% had used cocaine in their lifetime. Average use of marijuana was 2.3 times and cocaine 2.7 times in the last year (Table 2).

Drinking Practices

The only alcohol drink that a cantinera consumes at work is beer (Fernández-Esquer, 2003). On average, cantineras reported drinking on 15 working days in the last month. The highest number of drinks they consumed in a single occasion was 21 beers (see top row of Table 2). Cantineras also reported drinking an average of 11 beers on a “normal” day at work. While we did not use standard drinks, we have made a conservative estimate based on the 8 oz. “ponies” commonly described as part of their drinking routines (Fernández-Esquer, 2003). Based on these estimates, we converted self-reported drinking to standard drinks. On a normal day, they drink over one 6-pack (7.3 beers), and on a heavy drinking night, they drink over two 6-packs (14 beers) of standard 12-ounce beers. Even when not every night at work is a “good [earnings] night,” hostesses make an average of 28 “tickets” or the equivalent of US$224 on a good night. This estimate is based on an US$8.00 sales commission per beer, reported in the ancillary qualitative study (Fernández-Esquer, 2003).

Drinking Practices by Demographic, Work, and Risk Behavior Characteristics

Bivariate correlations are presented in Table 2. Demographic characteristics: Results indicate that younger women, those of more recent arrival in the United States and those reporting a higher monthly income were significantly more likely to report drinking more beers per single night, more beers in a normal day at work, and more tickets in a good night. Social network variables indicate that women who reported drinking more at work were more likely to report having fewer relatives and friends to trust. Cantina work practices correlations indicate that length of the work week was associated with higher frequency of drinking in the last month and with the number of tickets made in a good night. Type of waitressing job (higher score indicates higher risk of working for tickets) was significantly associated with all the drinking variables. Cantineras working for tickets were more likely to report a higher frequency and volume of drinking at work. Substance use variables correlations indicate that women who reported smoking more cigarettes per day were also more likely to report drinking more beers per single occasion. Additionally, more frequent marijuana and cocaine use in the last year was associated with reporting more tickets on a good night at work. Sex risk behavior variables that were significantly associated with all the drinking variables included having sex after drinking and reporting more sexual partners in the last year. In every instance, higher frequency and volume of drinking were associated with having sex after drinking more frequently and with reporting more sexual partners in the last year. Meeting one’s sexual partners in a cantina was significantly associated with normal day drinking and with making more tickers in a good night. Trading sex for money was significantly associated only with making more tickets in a good night.

Multiple Regressions

In order to identify the common predictors of cantineras’ drinking practices, we regressed the drinking variables on sociodemographic, work history, substance use, and sex risk variables. As the multiple regressions results resemble patterns described in the previous section, we provide a summary below of the most salient findings.

Frequency and Quantity of Alcohol Use

Social network, work practices, and sexual risk-taking variables significantly explained variance in frequency (adjusted R² = .16) and quantity (adjusted R² = .21) of drinking at work. Work practices were more efficient predictors of quantity than frequency of drinking at work, as the job type, years of work, and the total number of cantinas worked in the last 3 years explained variance in drinking quantity. By contrast, only length of the work week (days a week working in cantina) and the job type explained drinking frequency. Sexual risk-taking behaviors (having sex after drinking and number of sex partners in the last year) and having fewer trusted friends or relatives were associated with both greater frequency and higher quantity of alcohol consumption at work (Table 3).

Normal Day Drinking

The significant predictors of beers consumed in a normal day at work (adjusted R² = .16) included years working in a cantina, more frequent cocaine and less frequent marihuana use in the last year, and having sex after drinking. None of the sociodemographic or social network variables significantly predicted normal day drinking.

“Tickets” in a Good Night at Work

Sociodemographic, social network, work practices, and substance use predictors significantly explained variance in reported “tickets” (adjusted R² = .25). Being younger, more recently arrived to the United States, and making a higher monthly income were associated with more reported “tickets.” Working practices (years of work and the total number of cantinas worked in the last 3 years), having fewer trusted relatives, and smoking more cigarettes per day and less marijuana in the last year also predicted more “tickets.”²

Discussion

Overall, the results of the study indicate that cantineras participate in very hazardous drinking as evidenced by the frequency and volume of their drinking on the job. This drinking as a job practice is more common among hostesses, as it is required by their employment, but it is also reported by waitresses who drink socially (not paid for drinking) with customers. Results indicate that younger, more recently arrived cantineras with more experience working in cantinas drink higher quantities of alcohol on a routine weekly basis. The drinking of cantineras is not only associated with their demographic background, it is also related to their lack of social integration, as evidenced by the few friends and relatives they trust. It is possible that their social isolation may contribute to their heavy drinking, but it is also possible that their jobs and their drinking practices isolate them from the local Latino community, which disapproves of both.

The more a cantinera drinks, the more money she makes, although higher monthly income is only a significant predictor of the drinking of hostesses (i.e., tickets in a good night), an association that confirms the drinking for wages pattern that describes their work (Fernández-Esquer, 2003). Exposure to the heavy drinking context of cantinas is also associated with smoking more cigarettes per day, and with a higher likelihood of reporting recent use of cocaine. We suspect that the stimulant effect of cocaine may facilitate heavier drinking. By contrast, the negative association between heavy drinking and marijuana use may be due to the effects of this depressant drug, as it may interfere with the vigilance and alertness needed by a hostess trying to persuade customers to buy her beers, if she is to make many tickets in a work night.

If a cantinera drinks heavily, she is also more likely to report having sex after drinking and to report more sexual partners in the last year. Even though correlational analysis revealed this significant relationship, a striking finding of the regression analyses was the lack of association between sexual risk-taking and the drinking variables, particularly making “tickets” on a good night at work. This lack of significant association may indicate that drinking as a work practice is not necessarily related to sexual risk taking. However, it may also indicate that the relationship between sexual risk taking and the drinking variables could have been attenuated by cantineras’ underreporting of their sex-trading practices, as indicated in the “Results” section of this article.

Our results indicate that the predictors of drinking in a cantina are mostly associated with the occupational conditions under which cantineras operate. Their drinking at work does not represent a personal risk driven by personal needs and motives; rather, it is driven by the context of their work. It is also related to their status as recent immigrants who lack economic and social resources that allow them to obtain better jobs.

The lessons learned from the cantineras’ experience illustrate their hardships and also the influence of contextual risk factors that might be confronted not only by other cantineras but also by Latina workers in other occupations. Cantineras share with other Latina immigrants the hardships of their presence in the United States as undocumented workers with low education, limited English, and job skills. Their job opportunities are primarily in the service sector, and have in common similar low pay, long hours, and exploitive job conditions. The work of cantineras also overlaps with the work of women in the sex industry, as trading sex for money is an option that many cantineras take to supplement their income. The drinking arrangements at cantinas represent the exploitive trade of personal services in exchange for money, similar to those experienced by women in the sex industry. However, the job priorities of both groups are reversed: Sex workers trade sex for money as a primary transaction, while cantineras’ primary duty is to get paid for drinking beer with customers in the bar.

In the context of the United States, the drinking for wages pattern is a unique practice brought by Latino immigrants to the United States. The migration corridor across Texas most likely takes these practices to other communities where Latino immigrants are willing to participate in this practice: Latino men invest their money supporting the drinking practices of Latina cantineras who are willing to drink great quantities of beer and provide them with entertainment and company in exchange for money. Based on our observations and on the reports of other public health researchers and professionals, drinking in exchange for money exists in Latino enclaves in the Southwest, as well as other parts of the country. Additional studies are needed to develop interventions that address the social and economic needs of Latino immigrants and reduce high-risk drinking among Latina cantineras.

Limitations

This study represents an attempt to quantify the drinking practices of a group of Latina high-risk drinkers. As the primary goal of the parent study completed in 2001 was to test a cancer prevention program, the measures of drinking included in the original questionnaire did not include all relevant variables to fully capture this practice. For example, we did not document the size of the beer most commonly consumed in a cantina at the time of the study and it is possible that our estimated drinking volume based on “ponies” may be too conservative. In the 13 years since this project was completed, we have documented the existence of cantinas in other parts of Texas, and in 2008, we validated the continued existence of the drinking for wages pattern (Fernández-Esquer & Agoff, 2012). We also have firsthand reports that cantineras may have replaced the 8-ounce “ponies” for regular 12 oz. beers and suspect that their already hazardous drinking patterns may have worsened with the change in beer size. We acknowledge that data collected in 1998 may not represent current practices. However, based on the qualitative study we conducted in 2008 and the reports in the local press, we believe that the risky drinking of cantineras and its associated sex trade (Olsen, 2008) persist in the anti-immigrant climate currently widespread in the United States.

Drinking as an occupation is a practice that places cantineras at high risk of multiple health problems including alcohol abuse, sexually transmitted diseases, unwanted pregnancy, and intimate partner violence. In addition to these drinking-related risks, cantineras have increasingly become victims of sexual and human trafficking (Simons, 2012; Smith, 2014). As long as these Latina immigrants have limited opportunities to survive economically, this hazardous drinking practice will persist endangering their health, their relationships, their families, and their lives. We encourage researchers to address the needs of cantineras in other U.S. communities by creating programs that can reduce their heavy drinking and associated health risks, and their economic needs.

Footnotes

Acknowledgements

We would like to thank Craig Field, PhD, for his valuable editorial comments to this article.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by a grant from the National Cancer Institute (NIH Grant R29-CA-68113).

Notes

Authors Biographies

Maria Eugenia Fernandez-Esquer completed her PhD in environmental psychology at the University of Arizona. She is an associate professor at the University of Texas Health Sciences Center at Houston in the School of Public Health in the Division of Health Promotion/Behavioral Sciences. Her research explores the health risk behaviors of immigrant workers and marginalized Latino populations. She teaches courses on introductory and advanced health promotion theory and on multicultural perspectives in public health.

Pamela M. Diamond received her PhD from the University of Texas at Austin in educational psychology with a specialization in quantitative methods. She is currently at the University of Texas Health Sciences Center at Houston in the School of Public Health where she has a joint appointment in the Divisions of Health Promotion/Behavioral Sciences and Biostatistics. Her research focuses on health disparities among vulnerable populations such as prison inmates, abused elders, and recently immigrated Latinos. She teaches courses in applied statistics, latent variable models, and program evaluation.

References

Ayala

Carrier

Magaña

(1996). The underground world of Latina sex workers in cantinas. In Mishra

S. I.

Conner

R. F.

Magaña

J. R.

(Eds.), AIDS crossing borders: The spread of HIV among migrant Latinos (pp. 95-112). Boulder, CO: Westview Press.

Caetano

(1994). Drinking and alcohol-related problems among minority women. Alcohol Health & Research World, 18, 233-241.

Canino

(1994). Alcohol use and misuse among Hispanic women: Selected factors, processes, and studies. The International Journal of the Addictions, 29, 1083-1100.

Centers for Disease Control and Prevention. (1991). Behavioral Risk Factor Surveillance System Survey Questionnaire. Atlanta, GA: U.S. Department of Health and Human Services.

Farabee

Wallisch

Maxwell

(1995). Substance use among Texas Hispanics and Non-Hispanics: Who’s using, who’s not, and why. Hispanic Journal of Behavioral Sciences, 17, 523-537.

Fernández-Esquer

M. E.

(2003). Drinking for wages: Obligatory alcohol use among Cantineras. Journal of Studies on Alcohol, 64, 160-166.

Fernández-Esquer

M. E.

Agoff

M. C.

(2012). Drinking and working in a cantina: Misrecognition and the threat of stigma. Culture, Health & Sexuality, 14, 407-420.

Fernández-Esquer

M. E.

Cardenas-Turranzas

M. G.

(2004). Cervical cancer screening among Latinas recently immigrated to the United States. Preventive Medicine, 38, 529-535.

Fernández-Esquer

M. E.

Diamond

P. M.

Atkinson

J. A.

(2010). Lazos que Hatan: The influence of normative gender beliefs on the sexual risk behaviors of Latino men and women. Journal of Applied Social Psychology, 40, 2969-2994.

10.

Ferreira-Pinto

J. B.

Ramos

R. L.

Shedlin

M. G.

(1996). Migrant males and female sex workers: HIV/AIDS infection in the US-Mexico border. In Mishra

S. I.

Conner

R. F.

Magaña

J. R.

(Eds.), AIDS crossing borders: The spread of HIV among migrant Latinos (pp. 113-136). Boulder, CO: Westview Press.

11.

Finch

B. K.

(2001). Nation of origin, gender, and neighborhood differences in past-year substance use among Hispanics and Non-Hispanic Whites. Hispanic Journal of Behavioral Sciences, 23, 88-101.

12.

Hernández

M. T.

Sanchez

M. A.

Ayala

Magis-Rodríguez

Ruiz

J. D.

Samuel

M. C.

Lemp

G. F.

(2009). Methamphetamine and cocaine use among Mexican migrants in California: The California-Mexico epidemiological surveillance pilot. AIDS Education and Prevention, 21(Suppl. B), 34-44.

13.

Herrera

Campero

(2002). La vulnerabilidad e invisibilidad de las mujeres ante el VIH/SIDA: Constantes y cambios en el tema [Vulnerability and invisibility of women confronting HIV/AIDS: Constants and changes on the theme]. Salud Pública de México, 44, 554-564.

14.

Hondagneu-Sotelo

Avila

(1997). I’m here, but I’m there’: The meanings of Latina transnational motherhood. Gender & Society 11(5): 548-571.

15.

Hewitt

(2010, August 29). Cantinas in Harris officers’ crosshairs. Houston Chronicle. Available from http://www.chron.com

16.

Kjaerheim

Mykletun

Aasland

O. G.

Haldorsen

Andersen

(1995). Heavy drinking in the restaurant business: The role of social modeling and structural factors of the work place. Addiction, 90, 1487-1495.

17.

McAlister

A. L.

Ramirez

A. G.

Amezcua

Pulley

L. V.

Stern

M. P.

Mercado

(1992). Smoking cessation in Texas-Mexico border communities: A quasi-experimental panel study. American Journal of Health Promotion, 6(4), 274-279.

18.

Moore

R. S.

Cunradi

C. B.

Duke

M. R.

Ames

G. M.

(2009). Dimensions of problem drinking among young adult restaurant workers. American Journal of Drug and Alcohol Abuse, 35, 329-333.

19.

Norris

A. H.

Kitali

A. J.

Worby

(2009). Alcohol and transactional sex: How risky is the mix? Social Science & Medicine, 69, 1167-1176.

20.

Olsen

(2008, September 7). Despite crackdown, Houston cantina sex trade persists. Houston Chronicle. Available from http://www.chron.com

21.

Parker

K. D.

Weaver

(1995). Predictors of alcohol and drug use: A multi-ethnic comparison. Journal of Social Psychology, 135, 581-591.

22.

Simmons

(2012, December 30). The cantinera exposes human trafficking niche. Your Houston News. Available from http://www.yourhoustonnews.com

23.

Smith

(2014, May 20). Human trafficking sting named after disco hit nets 18 arrests at Houston cantinas. [Houston Press blog]. Available from http://blogs.houstonpress.com

24.

Treno

A. J.

Alaniz

M. L.

Gruenwald

P. J.

(1999). Drinking among U.S. Hispanics: A multivariate analysis of alcohol consumption patterns. Hispanic Journal of Behavioral Sciences, 21, 405-420.

25.

Umberson

Crosnoe

Reczek

(2011). Social relationships and health behavior across life course. Annual Review of Sociology, 36, 139-157.

26.

Wasserman

J. J.

Kutner

M. H.

(1983). Applied linear regression models. Homewood, IL: Richard D. Irving.

27.

Zemore

(2005). Re-examining whether and why acculturation relates to drinking outcomes in a rigorous, national survey of Latinos. Alcoholism: Clinical & Experimental Research, 29, 2144-2153.

28.

Zemore

(2007). Acculturation and alcohol among Latino adults in the United States: A comprehensive review. Alcoholism: Clinical & Experimental Research, 31, 1968-1990.