Introduction

The present study examines gender role portrayals in Indian television advertising using content analytical methodology developed by previous researchers in the field. It addresses both the frequency of appearance of men and women and the ways in which they are portrayed; in addition, by using the framework for analysis and the methodology developed by previous researchers in the field (e.g., Gilly 1988; McArthur and Resko 1975; Milner and Higgs 2004), the study aims to add to the existing knowledge base on the topic of gender role stereotyping in ads and to situate the findings in the context of those from other Asian nations.

Advertising has a major influence on our lives and plays a significant role in creating and disseminating a number of stereotypes including gender stereotypes (Fowles 1996; Jhally 1987; Leiss et al. 1986). As Dates and Barlow (1990) note, media stereotypes (including those present in advertising), are not harmless products of popular culture but are socially constructed images that are often one-dimensional and distorted in their portrayal of minority groups and women. The manner in which men and women are portrayed in advertising may also serve to maintain and reinforce existing stereotypes of men and women (Allan and Coltrane 1996; Furnham and Mak 1999; Manstead and McCulloch 1981). In addition, it is a critical agent of socialization and influences the way adults and children view themselves and learn appropriate gender behaviors (Frueh and McGhee 1975; Signorielli 1989, 1993). For example, social learning theory suggests that media images might lead to the imitation of depicted behaviors by males and females and to the creation of norms of acceptable behavior (Bailey 2006; Morgan and Signorielli 1990). In addition, such images may also be stored in memory and serve as cognitive scripts for later retrieval and use by others (Bailey 2006; Larson 2001). Cultivation theory suggests that consumers’ perceptions of social reality are heavily influenced by how they see themselves and others portrayed in the media (e.g., Shrum 1996; Tapper 1995). Many researchers (e.g., Gunter 1995; Manstead and McCulloch 1981) are of the opinion that television ads present idealized (rather than typical) images of appropriate behaviors and roles for men and women, making television advertising an important factor in developing notions of what appropriate behaviors and roles are for each sex.

The important role played by advertising in creating, shaping and reinforcing gender stereotypes has led to considerable interest in examining gender stereotyping in advertising. The influence of television advertising is likely to be even more significant in the case of a country like India where television was introduced only in the early 1970s than in countries where television was introduced much earlier (De Souza 2006). Further, the Indian television scene has changed dramatically over the past decade with increased domestic and international programming leading to increased interest in television programs (Bowman 2004; De Souza 2006).

Previous studies on gender stereotyping in television commercials indicate that gender differences are very prevalent in television ads. Such stereotyping has been found in ads from many different countries including several from Asia—for example, Japan (e.g., Arima 2003), Korea (An and Kim 2007; Kim and Lowry 2005), Malaysia (e.g., Bresnahan et al. 2001), Indonesia (Furnham et al. 2000b), Hong Kong (Furnham and Chan 2004; Furnham et al. 2000b), and Singapore (Siu and Au 1997; Tan et al. 2002; Wee et al. 1995). However, there have been very few studies on gender roles in Indian television ads or on ads from other countries in South Asia (i.e., the Indian subcontinent). As a market, India is an extremely attractive one to multinationals for “India is at an exciting tipping point in its socio-economic progress that makes it the cynosure of global investors” (Reddy et al. 2007; p.3). India is the fourth largest economy in the world with a sizeable middle class of over 300 million people; it also has the third largest television viewing market in the world (Bowman 2004; Forrester 2007). These factors make understanding gender role portrayals in this country is important to both local and multinational companies.

While there are strong similarities between India and other Asian countries (Ahmed 1998; Perlow and Weeks 2002), there are also significant differences in terms of values, religion, and other factors and hence, findings from other Asian nations may not be entirely generalizable to the Indian context. For example, Hofstede (2001) found that while like most Asian nations, India scored high on masculinity and power-distance, it differed on other cultural dimensions (long-term orientation, uncertainty avoidance and individuality). In addition, India is different from other Asian nations in terms of religion as it is the only country in the world where Hinduism is the predominant (80%) religion. Religion has a significant impact on any culture; for example, Hofstede (2001) found that East Asian countries where the majority follows Buddhism, Taoism, or Confucianism have very different values than predominantly Muslim countries such as Pakistan. Further, as Kumar (1994) and Munshi (1998) note, South Asia has its own discourses on gender and power, shaped to a great extent by its colonial history and the significant socio-economic changes that have occurred within the region after the end of colonialism. In India in particular, there has been a growing trend towards portraying women as liberated yet traditional, as persons who still bear the responsibility of nurturing and caring for others—in other words, there is an indigenously developed sense of modernity with corresponding notions of the ‘New Indian Women’ (Munshi 1998). Finally, as Wolf (2009) notes, feminism and notions of gender equality in the Indian context have evolved very differently (than in western nations); equality is seen in India not in individual-centered but in family-centered ways and claims to equality do not negate women having valued roles in the home, or prizing family above all. This, in turn, may influence how marketers portray men and women in advertisements.

Over the years, television has become an important medium in Indian society. The number of television channels has grown from six in 1991 to more than 230 in 2006 (Bowman 2004). The Indian TV industry is expected to grow from $4.3 billion in 2006 to $9.5 billion in 2010 and overtake print media in size (De Souza 2006). All of the above indicate that the Indian television scene is changing and growing, and the impact of television on Indian society is increasing. Yet, to date, few researchers have examined gender role portrayals in Indian television ads. An extensive literature review yielded only a few journal articles on the topic. Of the four journal articles found on gender role portrayals in Indian television ads, three were from the 1990s. The only article (Jha Dang and Vohra 2005) published in the past five years indicated that while men and women were portrayed in stereotypical ways in Indian television ads, the portrayals differ from those found in western nations. This study, while a good starting point, focused only on a few aspects of gender stereotyping (frequency of appearance, products appeared for, age, roles, and settings), had a relatively small sample and did not examine variables such as use of voiceovers (more details of this study are provided in the next section).

To summarize, India is a large and growing market for television advertisers and there is very little research on gender role portrayals in television ads in India; in addition, there are differences between India and other nations in terms of cultural values and as such, it a study on gender roles in India is perhaps necessary and warranted as findings from other nations (on gender role portrayals in television ads) may not be totally generalizable to the Indian context. This study will examine gender role portrayals in Indian television ads from two perspectives: the frequency of appearance and the nature of portrayals. The results will be compared with those from other nations—especially Asian countries. Chi-square tests will be the primary method used for hypothesis testing.

Gender Role Portrayals in Television Ads

Bardwick and Schumann (1967), in one of the earliest studies on gender role portrayals in television ads found that women (compared to men) appeared more often as homebound and as housewives in television ads. This study was followed by several others that looked at gender stereotyping in television ads; the most notable among these are the works of Gilly (1988) and McArthur and Resko (1975). These researchers developed a well-defined research methodology and clear definitions of key variables which enabled others to use the same framework and thus compare gender stereotyping across countries.

Most of the research on this topic has focused on gender stereotyping in terms of two related but distinct aspects: the frequency of portrayals and the nature of portrayals. Results indicate that men are more likely than women to appear in ads in general and in ads for automotive, technical and financial services (Furnham et al. 2000b; Furnham and Voli 1989; Neto and Pinto 1998). However, contradictory findings have been reported by some researchers (e.g., Craig 1992; Nassif and Gunter 2008; Uray and Burnaz 2003). Women, on the other hand, have been found to appear more often in ads for domestic products (e.g., Allan and Coltrane 1996; Ganahal et al. 2003; Kim and Lowry 2005) and in those for body or personal hygiene products (e.g., Shrikhande 2003; Valls-Fernandez and Martinez-Vicente 2007). Similar findings—in terms of overall frequency of appearance and appearance for specific product types—have been found in the Indian context in the case of magazine ads (Das 2000).

As noted by Furnham and Bitar (1993), ads use voiceovers for conveying important information about a product in an authoritative manner. Previous studies have consistently indicated that male voiceovers are more likely to be used than female voiceovers in the United States, Britain, France, Spain and many other countries (Bresnahan et al. 2001; Bretl and Cantor 1988; Ferrante et al. 1988; Furnham and Bitar 1993; Furnham et al. 2000a; Milner and Higgs 2004).

In terms of the nature of portrayals in ads, men and women seem to differ in the following: the roles they are shown in, their demographic characteristics (e.g., age, employment status, etc), their source credibility, the arguments used by them to persuade consumers, and the settings that they appear in. Unlike men, women are more likely to be portrayed in dependent roles (e.g., mother, spouse) than independent roles such as narrators, employees and/or celebrities (e.g., Furnham et al. 2000b), and in home rather than business, outdoor or public settings (e.g., Milner and Collins 2000; Shrikhande 2003; Valls-Fernández and Martinez-Vincente 2007). In addition, women’s credibility in ads seems to originate from their role as users (rather than experts or authority figures) of the product being advertised (e.g., Furnham et al. 2000b; McArthur and Resko 1975; Neto and Pinto 1998).

In terms of demographic variables, past research indicates that female characters in ads are more likely to be younger than male characters. This finding has been consistent across several countries (e.g., Spain: Valls-Fernandez and Martinez-Vincente 2007; Australia: Mazella et al. 1992; Milner and Higgs 2004; Japan: Arima 2003). In addition, women in television commercials are less likely than are men are to be shown in working roles (e.g., Wiles and Tjernlund 1991).

The above findings have been found to hold true in almost all countries including many western nations (e.g., UK: Furnham and Bitar 1993; Furnham and Farrager 2000; Manstead and McCulloch 1981; Italy: Furnham and Voli 1989; Spain: Valls-Fernández and Martínez-Vicente 2007; Portugal: Neto and Pinto 1998; France: Furnham et al. 2000a), and Asian countries (e.g., Japan: Arima 2003; China: Cheng 1997; Hong Kong: Furnham et al. 2000b; South Korea: Kim and Lowry 2005, 2007; Malaysia: Bresnahan et al. 2001; Indonesia: Furnham et al. 2000b; Singapore: Siu and Au 1997; Tan, et al., 2002), and African nations (e.g., Zimbabwe: Furnham et al. 2001; Kenya: Mwangi 1996). Thus, gender stereotyping of men and women in television ads seems to be a worldwide phenomenon.

While the general trends in gender portrayals are similar, it should be noted that the exact nature of gender stereotyping does seem to vary slightly across cultures. For example, An and Kim (2007) examined differences in gender role portrayals in ads between the US (a country that scores high on Hofstede’s masculinity dimension) and South Korea (a country that scores much lower on masculinity). Results indicate that South Korean ads featured characters in relationship themes more often than ads from the US; women were also more likely to be the main characters in South Korean ads than in those from the US (An and Kim 2007). Similar results have been found in studies comparing ads from Japan and the US (Sengupta 1995), and Sweden and the US (Wiles and Tjernlund 1991; Wiles et al. 1995).

Gender Role Portrayals in Indian Ads

Past research indicates that in India, as in other nations, men and women are likely to be shown in stereotypical ways in television ads (Gupta and Jain 1998; Jha Dang and Vohra 2005; Munshi 1998) and in print ads (Das 2000; Nelson and Paek 2005). Munshi (1998) examined Indian television ads in the 1990s and found that women were portrayed primarily as housewives. Gupta and Jain (1998) also found that Indian television ads served to reinforce the traditional and stereotypical images of women as subordinate to men. The most detailed study on gender role portrayals in Indian television ads by Jha Dang and Vohra (2005) confirms the above findings. Jha Dang and Vohra (2005) found that compared to men, women appeared more often for household products and personal or beauty products and were less likely to be shown as being employed than were men. Similar findings have been found in magazine ads. For example, Das (2000) found that men appeared more often in ads for automotive, financial services, and electronic products than did women but that women appeared more often for home and beauty products than did men.

However, slight differences between Indian ads and ads from other nations have also been found. Jha Dang and Vohra (2005) found that (a) women appeared more often in television ads than men, (b) there was no significant difference in the age of male and female characters, (c) when the characters were shown as employed, the proportion of women shown in occupations such as doctors, engineers or architects was similar to that of men, and (d) the proportion of women shown as sex objects or in purely decorative roles in Indian ads was lower than that in western ads. Das (2000) in a study of Indian magazine ads also found such differences between western and Indian ads. For example, the percentage of females portrayed as sex objects in Indian ads was lower than that in ads from western nations (Das 2000). These differences might be due to the indigenously developed sense of modernity in Indian culture (Munshi 1998), the greater emphasis on family values (Wolf 2009), and/or the unique religious (i.e., predominantly Hindu) values that permeate Indian society. For instance, the greater emphasis on family values might have led to the lower proportion of female portrayals in purely decorative roles compared to that in ads from other nations and the indigenously developed sense of modernity in Indian culture may have led to the willingness to show women who were employed as being in equally high-status occupations as men who were employed.

To summarize, the literature review indicates that (a) across cultures, men and women are generally portrayed in stereotypical ways; (b) there are slight differences in gender role portrayals across cultures; (c) Indian ads follow the global trends in many ways with women (compared to men) being portrayed more often as younger, in dependent roles, in home settings, and in ads for home and beauty products ; and (d) Indian ads differ from ads from other (especially western) nations in terms of women’s frequency of appearance in decorative roles and their job status (when shown as employed, women and men appear in jobs of equal status).

Examining gender stereotypes in Indian television ads will provide an understanding of the existing gender stereotypes in Indian society. In addition, such a study can also be of practical relevance to marketers. Given the increasing participation of Indian women in the workforce and their influence on purchase decisions, it is important for marketers to understand existing gender portrayals and to adapt them to the changing Indian culture (Reddy et al. 2007). As Reddy et al. (2007) note, today’s Indian women express multi-faceted behavior, are no longer limited to their traditional domestic roles (i.e., mother, wife, daughter, etc), and are increasingly active in other settings. For example, 72% of the teenage girls in major cities state that they would like to have a career (Reddy et al. 2007). Hence, portraying women as influential and powerful by showing them as employed and/or as experts and by using female voiceovers may all be necessary to target the Indian female market (Kotwal and Sahni 2008). Understanding existing role portrayals is a necessary first step to assessing their relevance to today’s Indian society and modifying them if necessary.

Research Hypotheses

The major objective of this study was to examine gender role portrayals in Indian television ads and to compare them with those found in other countries. The research methodology and format established by previous researchers in the field was followed as closely as possible to facilitate the comparison of gender role portrayals in Indian television ads with those found in other nations. The intent of this study was to choose ads that can cumulatively have a great impact on the Indian television viewer.

Based on the literature review, the following hypotheses were developed:

  • H1: There will be more men in Indian television ads than women.

  • H2: Men will be used more often for voiceovers than women.

  • H3: Women will appear more often in ads for home and beauty products while men will appear more often in ads for automotive, electrical/electronic equipment and financial services.

  • H4: Women will be portrayed as younger compared to men.

  • H5: Women are less likely to be portrayed as employed than male characters.

  • H6: Women are more likely to be presented in relationship roles (i.e., as parent, spouse, child, or other relationship) than men.

  • H7: Women are more likely to appear in home settings than men.

  • H8: Women will be shown more often as product users while men will be shown more often as authority/expert figures.

Past research findings from other nations and India (cited earlier) provide the rationale for hypotheses H3, H5, H6 and H7. Hypotheses H2 and H8 have been found to hold true in across all cultures, but have not been tested in the Indian context. Finally, H1 (overall frequency of appearance) and H4 (age of characters) are ones for which past research has yielded contradictory findings. For example, while most researchers have found that men appear more often in ads than women, contradictory findings have been found in certain studies (e.g., Coltrane and Adams 1997; Bresnahan et al. 2001). In India too, the findings have been somewhat contradictory; while Das (2000) found that men are more likely to appear in magazine ads than women, Jha Dang and Vohra (2005) found the opposite to be true in Indian television ads. But, given that the majority of research supports the finding that men appear more often in ads than women, H1 was formulated in its current form—i.e., that men will appear more often in Indian television ads than women. Similarly, it has been hypothesized that compared to men, women in Indian television ads will be portrayed as younger (H4) as most researchers (in a variety of countries; cited earlier) have found this to be true. In India, the only study on television ads (Jha Dang and Vohra 2005) did not find a statistically significant difference in terms of the age of characters, but the direction of the difference was the same as that found in other countries (i.e., women tended to be younger than men).

Method

Sample

Given the language diversity in India, ads in three different languages—English, Hindi, and Tamil—were included in the study. English is a national language in India, but is spoken more commonly in major cities and towns and by people of higher socio-economic level. Ads in English tend to be from larger, national/international firms. Compared to programs in English, Indian languages programs cater to different target audiences and are used more often by local firms than by international ones. Given these differences, ads in two Indian languages—Hindi and Tamil—were also examined. Hindi (a North Indian language and the Official Union Language), is spoken by more people (over 35% of the population) than any other language in the country (Government of India 2009); Tamil, a south Indian language, is spoken in the state of Tamil Nadu. Programs in regional Indian languages also tend to have a higher number of ads from smaller (often local) companies than either English or Hindi. Finally, the fact that the researcher is fluent in Tamil and Hindi was also a major factor in the choice of these languages.

Ads were chosen from three types of channels: four national channels (the public television channel Doordarshan, NDTV, Zee, and Star), four regional Tamil channels (Sun, Jaya, Raj, and the regional version of Doordarshan), and four international channels (V, MTV, Cartoon Network, and AXN). The four regional channels broadcast Tamil programs and hence most (over 92%) of the ads are also in Tamil; the national and international channels broadcast programs in English and Hindi and most of the ads (over 97%) in these channels are in English or Hindi. Thus, there is a high correlation between the type of channel and the language of the ads. The channels were chosen based on audience share figures and were the top ones in national/regional audience shares (Nair 2003; Savage 2004). While the sample was limited to the four most popular channels in each category, it is felt that this sample would be representative of the total population of Indian television ads—in the chosen time period—for the following reasons: (1) unlike in North America, Indian television channels do not specialize by type of programs in any significant way (i.e., there are no Indian equivalents to the Women’s Network, History Channel, CNN, or channels specializing in cooking shows, weather information, travel, etc) and hence, the omission of ads from less popular (but not specialized) channels is less of a problem in the Indian context; (2) while there might be differences in the type of people watching specific programs, the random selection of time slots and days of the week should ensure that the sample would include most audience types; and (3) together, these channels account for most of the Indian television audience—thus ensuring a representative sample of ads watched by Indian television viewers.

Even assuming that the sample is not perfectly representative of the total population of Indian ads, cultivation theory suggests that the impact media portrayals is likely to increase as exposure to such portrayals increases (e.g., Shrum 1996; Tapper 1995). If, as cultivation theory suggests, certain themes are common across programs and viewers’ perceptions of reality are molded by cumulative exposure to media and by the ‘mainstreaming’ of media images (Gerbner and Gross 1976; Gerbner et al. 2002), then television ads in popular programs have to have greater impact on the gender role perceptions of Indian television viewers than the ads in less popular programs; choosing a random sample of television ads from the most popular television channels is thus justifiable given the focus of this study. It should also be noted that most of the studies on gender role portrayals in other nations have focused on primetime television or on programs shown on major networks and hence have been limited to television ads appearing in popular channels and/or shows and hence make the comparisons of the findings from this study to others relevant.

Ads were chosen during a 2 week period (including weekends) in October, 2004. (Note: It is recognized that choosing ads over a longer time period would have made the sample more representative of the total population of Indian television ads in any given year. The time frame was chosen due to practical constraints.) For each channel, two hours of programming were taped; each channel was taped for one hour during the day (9–5) and one hour during evening. The day and evening time slots were chosen randomly. Thus, each channel was taped for four hours during the two week period except for Sun TV. The number of ads per hour of programming was lower in Tamil language channels than in the national/international channels, and hence, the most popular Tamil channel—Sun TV—was sampled for an extra day to increase the sample of Tamil ads). Following the method used by previous researchers (e.g., Arima 2003; Furnham and Farrager 2000; McArthur and Resko 1975; Nassif and Gunter 2008; Uray and Burnaz 2003), repeat ads, institutional ads and ads with no adult human character were excluded from the study; movie promotions and channel self-promotions were also excluded. While some researchers include repeat ads (e.g., Gilly 1988), the exclusion of repeat has become common practice (please refer to the above studies), as most researchers feel that it is more important to examine the number of distinct portrayals of men and women than to include multiple copies of the same ad. In addition, the exclusion of repeat ads also makes comparisons with other studies feasible. Up to two main characters were coded in each ad. To be coded, a character had to appear for a minimum of 3 s visually or verbally. If more than two characters appeared in an ad, the two most important ones were chosen. As shown in Table 1, the final sample consisted of 318 ads (72 English, 151 Hindi and 95 Tamil ads) with 627 characters in total (157 from English ads, 297 from Hindi ads and 175 from Tamil ads).

Table 1 Variables included in the study.
Full size table

Variables

For each ad, the following ad-specific variables were coded: product type, language, and use of voiceover. Then, for each of the two main characters, the following were coded: sex, setting in which the character appeared, the mode of presentation (visual only or visual + speaking), age, employment status, role played, argument used, and type of credibility. The coding scheme used by previous researchers (especially Gilly 1988; McArthur and Resko 1975; Milner and Higgs 2004) was followed, with very slight modifications. For example, product type had to be modified to suit the Indian context; as there were several ads for clothing, this category was subdivided into clothing for males, clothing for females. Table 1 provides details of the coding scheme.

Coding

Four coders (two men and two women), all fluent in all three languages were trained in content analysis. They were given a written description of the variables and underwent a 2-h session during which the variables were clarified; later, they went through 45 ads for training purposes; each coder coded these ads (which were not part of the study) and differences between them were resolved through discussion (with the researcher being present to provide clarifications). By the end of this session, the intercoder agreement exceeded 85% for all variables. Given the difficulty in finding research assistants fluent in all three languages and the time constraints faced by the RAs, after this, the ads were divided into two sets and a team of two coders was assigned to code each set. Several other researchers (e.g., Milner and Higgs 2004; Uray and Burnaz 2003) also use the one male–one female team system for coding. The team members had to first agree upon the two most important characters, then code the variables in the study independently, and finally, meet and come to an agreement on the coding. When team members could not agree on the coding of an ad or the two most important characters, they contacted the researcher. It should be noted that except for a few instances, the teams were able to sort out the differences themselves. When differences arose between them, they approached the researcher for further clarifications or a tie-break.

Coding Reliability

5% of the ads were coded by all coders to check for intercoder reliability and Scott’s pi. As shown in Table 1, intercoder reliability figures were more than satisfactory for all variables; the overall interrater agreement was 91% across all variables and Scott’s pi was .77 indicating that the overall coding reliability was above the acceptable norms. The intercoder unitizing reliability was 1.00 indicating that the raters were able to agree on the two most important characters in these ads. As Table 1 indicates, the Scotts’s pi for age was below acceptable norms and hence the results of the chi-square tests for H4 (the ratio of women portrayed as younger would be higher than that of men) are to be treated with caution.

Data Analysis

Data were analyzed using SPSS. All hypotheses were tested using chi-square tests. As Schwab (2004) notes, while chi-square tests may indicate that there is a statistically significant difference between groups, it is difficult to understand which one of the cells led to the significant result without examining the standardized residuals. Hence, post-hoc tests were conducted to examine the size and direction (positive/negative) of the standardized residuals. These values were used to identify the cells that were producing the statistically significant chi-square results. The significance level of the standardized residuals in post-hoc chi-square tests indicate the variables that led to a significant result in a chi-square test while direction of the standardized residual (positive or negative) indicates if the actual frequency in the sample was higher or lower (i.e., whether there is overrepresentation or underrepresentation) than what could be expected (Schwab 2004). (Note: When the results of post-hoc chi-squares are discussed in the next section, the terms ‘overrepresented’ and ‘underrepresented’ are used to denote that the actual frequency in the sample is above/below the expected frequency; to reduce Type 1 error, the cutoffs for significance levels was set at p = .05 or higher.)

Results

H1: Men will Appear More Often than Women in Indian TV Ads

Of the 627 characters coded, 271 (43.2%) were females and 356 (56.8%) were males, χ2 (1; N = 627) = 11.52, p < .001; as there were significantly more male than female characters in the ads, the first hypothesis is supported (Table 2).

Table 2 Differences in frequency of appearance, voiceovers, and products appeared for.
Full size table

H2: Male Voiceovers will be Used More Often than Female Voiceovers

Chi-square tests support the hypothesis that male voiceovers are more likely to be used than female voiceovers in Indian television ads, χ2 (2; N = 278.62), p < .001. Of the 318 ads, 280 had adult voiceovers; female voiceovers were used in only 28 of the 280 ads (or 10%), while male voiceovers were used in 225 (80.3%) ads; 27 ads (9.7%) used both male and female voiceovers (Table 2).

H3: Men will Appear for Technical and Financial Products While Women will Appear for Home and Beauty Care Products

Since products were classified into ten categories, an overall chi-square test for all categories combined and individual chi-square tests for each product type were carried out. The overall chi-square test was significant, χ2 (9; N = 624) = 37.92, p .001. Of the female portrayals, the highest proportions were for body (20.7%), food (19.2%), and home-related products (12.9%); in the case of male portrayals, the highest proportions were for food items (16.3%) and electronic (15.4%) products. Men also appeared frequently for body products (10.7%), home item (10.4%), and services (10.4%). However, as shown in Table 2, the differences between men and women were only significant for five of the ten product types (body, automobiles, neutral clothing/watches/jewelry, electronic appliances, and ‘other’). Thus, H3 seems to be supported in the case of women as they appear more often for beauty, food and home products, but not in the case of men; hence, this hypothesis is only partially supported.

H4: Women will be Portrayed as Younger than Men

Female characters in Indian television ads were younger than their male counterparts, χ2 (2, N = 48.75), p < .001 (Table 3). 86% of the female characters were under 35 years of age, while only 61.8% of males fell into the same age category. As shown in Table 3, the standardized residuals in post-hoc tests confirmed that women were underrepresented (i.e., the actual frequency in the sample was lower than what was expected) in the older age categories and overrepresented (i.e., the actual frequency in the sample was higher than what was expected) in the younger age category in the actual sample compared to the expected frequencies. Thus, H4 is supported though it should be noted that the Scott’s pi for this variable was below acceptable norms and hence the results should be treated with caution.

Table 3 Demographic characteristics of Indian television characters.
Full size table

H5: Men will be Portrayed as Being Employed More Often than Women

The chi-square performed indicated that there was a significant difference in the employment status of male and female characters in Indian ads, χ2 (3, N = 627) = 43.63, p < .001. The results (Table 3) also indicate that in Indian ads, the proportion of women represented as having no clear occupational status lower than that of men, (80.4% vs. 60.7%). Women were less likely to be portrayed as employed (7%) than were men (23.9%). The standardized residuals in the post-hoc chi-square tests indicate that in the actual sample, the women were significantly overrepresented in the ‘unclear’ occupational category but underrepresented in the ‘employed’ category; men, on the other hand, were underrepresented in the ‘unclear’ category and overrepresented in the ‘employed’ category. The results support H5.

The data were further analyzed to examine if men and women in this sample of Indian ads were portrayed in different types of occupations. Chi-square tests indicated that there were significant differences in the type of jobs that men and women were portrayed in this sample; while both men and women were shown more often as celebrities than in any other occupation, men were more likely than were women to be shown as business or non business professionals, χ2 (3, N = 177) = 11.31, p < .05 (Table 3). The post-hoc chi-square tests did not yield any significant results.

H6: Women will Appear in Home Settings While Men will Appear in Work/Public Settings

Since only 10 ads (3.1%) were in an office setting, this category was combined with the ‘public’ category. The results are shown in Table 4. As hypothesized, women and men appeared in different settings in Indian television ads, χ2 (2, N = 627) = 25.72, p < .001. The standardized residuals in post-hoc tests indicated that in this sample, women were overrepresented and men were underrepresented in home settings compared to expected frequencies. These tests also indicated that compared to expected frequencies, women were underrepresented in outdoor settings.

Table 4 Other characteristics of Indian television characters.
Full size table

H7: Women will Appear More Often than Men in Relationship Roles

Due to low frequencies, the role categories had to be slightly modified before further analysis could be conducted. The roles of ‘child’ and ‘other relationship’ were grouped together as the numbers in each were not very high. Cases in which the role of the character was unclear (34.7% females vs. 31.7% males; not statistically significant) were omitted from further analysis. This left 420 characters. Of the characters portrayed in clear roles, that there were significant differences between male and female characters, χ2 (4, N = 420) = 50.06, p < .001 (Table 4). Standardized residuals in post-hoc tests indicated that compared to expected frequencies, there were more women in ‘parent’ and ‘spouse ‘categories and less in the ‘employed’ category. Men, on the other hand, were overrepresented in the ‘employed’ category and underrepresented in any relationship role than expected. Only 46.9% of the male characters were portrayed in relationship roles compared to 80% of the female characters. Thus, this hypothesis was also supported.

To make the results comparable with those from other studies (e.g., Furnham et al. 2000a), the “unclear” category combined with the “independent” category, and compared with all the relationship roles (parent, spouse, child, homemaker, other relationship). Still, women were far more likely to be portrayed in dependent or relationship roles than were men (52.4% vs. 32%; χ2 (1, N = 627) = 26.44; p < .001). As shown in Table 4, the post-hoc chi-square tests were also significant. The standardized residuals in these tests indicate that women were overrepresented in the relationship role while men were underrepresented in the same role; the reverse was true in the case of the ‘independent’ role category.

H8: Women’s Credibility will be Based on Status as Users While Men’s will be Based on Their Status as Experts

The credibility basis was clear in the case of 77.8% of the women and 71.6% of the men in these ads. As in previous studies, women were far more likely to be portrayed as users rather than as authority figures; the opposite was true in the case of men. In cases where the credibility basis was clear, 90% of the women were shown as users while only 83.9% of the men were portrayed in a similar manner; thus, while only 10% of the women were portrayed as authority figures, 16.1% of the men were portrayed as having some type of authority, χ2 (2, N = 627) = 7.05; p < .05. While the overall chi-square value for this test was significant, post-hoc tests indicated that no one category contributed significantly to the results (Table 4).

While there were no specific hypotheses related to language, the data were analyzed again with language being controlled for to examine if the gender roles portrayed in Indian ads varied by language. (Note: As noted earlier, language of the ads and channel type were highly correlated and hence differences by channel type were not analyzed). As shown in Table 5, in all languages, there were more men than women; the difference between the ratios of women to men was the least in English and the most in Tamil. Thus, the first hypothesis is supported in all languages. Similarly, the third hypothesis (voiceovers) and the seventh (employment status) were also supported in the case of all three language ads; male voiceovers were more popular than female voiceovers and women were less likely than were men to be shown as employed in all three languages. In the case of all other hypotheses, there were no differences in the portrayal of men and women in the English ads but in the case of ads in Tamil and Hindi, all other hypotheses were also supported. Thus overall, the results indicate that gender differences were less prevalent in English ads than in the two Indian language ads as only three of the ten hypotheses were supported in the case of English ads. All ten hypotheses were supported in the case of both Hindi and Tamil ads.

Table 5 Differences between English, Hindi and Tamil television ads (Only key details of significant findings are shown; the highest percentages are shown when there are more than two categories).
Full size table

Discussion

The results of this study indicate that gender stereotypes are prevalent in Indian television ads as compared to men, women are portrayed (1) less frequently as main characters or voiceovers, (2) more frequently in ads for female-oriented products, (3) more often in home settings, (4) more often in relationship roles, (5) less often as being employed, and (6) less often as authority figures. These findings are in line with findings from other nations, especially those from Asia and lend support to the findings of Jha Dang and Vohra (2005).

This study indicates that there are significant differences between Indian television ads based on language. This is probably due to the fact that English language ads cater to a more upscale, better-educated target market compared to those in Indian languages. English channels are only available through satellite or cable services which are more expensive options that are available more readily in urban settings than in rural areas. In addition, English ads appear in channels such as MTV and Cartoon Network which offer international programs that are often broadcast from other Asian nations. Thus, both the types of programs and the nature of the target market might have led to lower stereotyping in English ads than in Tamil or Hindi ads.

Comparison with Gender Role Portrayals in Western Ads

While the above findings are in line with recent findings from other nations, some interesting differences were also noted For example, the proportion of women in Indian television ads (43.7%) was lower than the proportion found in ads from most other countries (e.g., UK: Furnham and Farragher 2000; Nassif and Gunter 2008; Spain: Valls-Fernandez and Martinez-Vicente 2007; USA: Rizor 2003; Shrikhande 2003; France: Furnham et al. 2000b). Another difference between Indian television ads and ads from other nations was the use of voiceovers. In this study, only 10% of the ads used a female voiceover; this proportion is much lower than that found in Britain (Nassif and Gunter 2008), Spain (Valls-Fernandez and Martinez-Vicente 2007), or the United States (Kim and Lowry 2007).

As in the case of many western nations such as the US (e.g., Kim and Lowry 2007), Britain, (e.g., Nassif and Gunter 2008), France, Denmark (Furnham et al. 2000a), or Spain (Valls-Fernandez and Martinez-Vicente 2007), men and women appeared in ads for different products; the proportion of women appearing in ads for food, home, and beauty products was higher than that of men while the reverse was true in the case of ads for automobiles, financial services, and electronic items. However, this study found that women were also likely to appear for other products which made the proportion (52.8%) of women appearing in female-oriented products (e.g., home, food or beauty products) in this study lower than the proportion of women appearing for these products in ads from Britain (69.5%; Furnham and Farragher 2000), New Zealand (63.2% Furnham and Farragher 2000), France or Denmark (63.3% and 76.6% respectively; Furnham et al. 2000a).

In terms of occupation, while there is a gender difference in the predicted direction (i.e., the proportion of women portrayed as being employed was lower than that of men), if employed, the type of occupations that men and women held in Indian television ads were different from those found in most western nations such as the US, Sweden, and Denmark (e.g., Milner and Collins 2000). Compared to ratios of women nonbusiness professionals to business professionals in television ads from these nations, the ratio found in this sample of Indian ads was higher. Finally, in terms of their roles (relationship vs. independent), in this study too, men were more likely to be portrayed in independent roles than were women; however, significantly higher proportions of male and female characters were portrayed in relationship or dependent roles in India (females: 76.2%, males: 46.9%) than in France (Furnham et al. 2000a), New Zealand (Furnham and Farragher 2000), Britain, (Furnham and Farragher 2000) or the United States (Kim and Lowry 2007).

To summarize, the overall findings of this study are in line with those from western nations such as Britain, USA, France, Denmark, New Zealand, Australia and Denmark. But, the extent of stereotyping seems somewhat different. Women were much less likely appear in Indian ads or to be used as voiceovers than in ads from western nations, but were also less likely to be limited to beauty, home and food products than in ads from western nations. Finally, both males and females were more likely to be shown in relationship roles in Indian ads than in ads from most western nations.

Comparison with Ads from Other Asian Nations

One of the primary objectives of this study was to examine whether Indian television ads were different from those from other Asian nations in terms of gender stereotyping. Since no other study has examined gender roles in advertising from other South Asian countries (i.e., Pakistan, Bangladesh, Sri Lanka, Nepal), comparisons can only be made with ads from other Asian nations (e.g., Japan, Korea, Taiwan, Turkey, Singapore, and Malaysia).

While the findings from this study follow the general patterns found in other Asian countries, there are some interesting differences. For example, in terms of frequency of appearance, there were fewer women in Indian television ads than in ads from Japan (Arima 2003; Bresnahan et al. 2001), Korea (Kim and Lowry 2005), Singapore (Siu and Au 1997; Tan et al. 2002), Turkey (Uray and Burnaz 2003), Malaysia or Taiwan (Bresnahan et al. 2001). In all the above countries, women were more likely to appear in television ads than men but in the present study, women accounted for only 43.6% of the central characters. While male voiceovers were more popular in all countries, the percentage of female voiceovers in this study (10%) was much lower than that found in studies from Japan, Taiwan, or Malaysia (32%, 19%, and 19% respectively; all figures from Bresnahan et al. 2001); only Korea had a lower percentage (9.8%; Kim and Lowry 2005) of female voiceovers. Finally, in this study, men and women were found to appear in ads for all types of products while in studies from other Asian nations, men often did not appear for certain products (e.g., beauty care product ads in Singapore; Tan et al. 2002) and women did not appear for others (e.g., automobiles and CD ads in Singapore; telecommunication ads in Malaysia; Tan et al. 2002).

The present study indicates that while the overall nature of gender role portrayals in Indian television ads are similar to those found in other Asian countries, differences do exist. For example, both men and women were more likely to be portrayed in relationship roles in India than in ads from Korea, (30.8%; Kim and Lowry 2005) or Indonesia (7.5%: Furnham et al. 2000b). Parental and spousal roles were also more common in Indian ads than in ads from other Asian nations; 48% of women and 29.6% of men in this study were portrayed in one of the above roles; the figures for Korea for example, were 37.4% and 21.3% (Kim and Lowry 2005).

Indian ads also reflected the general trend found in other Asian countries (e.g., Korea: Kim and Lowry 2005; Turkey: Uray and Burnaz 2003) in terms of the employment status of the characters. Women in Indian ads were also less likely than were men to be portrayed as employed (7.0% vs. 23.9), or as celebrities (7.7% vs. 14.0%). But the percentage of women shown as business professionals was lower in this study (2.4%) than in Korean ads (An and Kim 2007) where it was over 10%. Finally, compared to Indian television ads, Korean ads were also more likely to use both men and women as experts rather than users (Kim and Lowry 2005).

In summary, while the overall results of this study are in line with those from most other countries, interesting differences were also found even when the results were compared to those from other Asian countries. Thus, the present study indicates that cultural contexts play a significant role in the nature and extent of gender stereotypes present in television ads and emphasizes the need for country-specific studies. Some of the differences may be attributable to demographic differences between India and other Asian nations. For example, India has a much higher proportion of people under 35 years of age (over 60%; United States Census Bureau 2004), than China, Japan, Indonesia, Hong Kong, Singapore, or Taiwan and this might explain the higher proportion of younger characters found in this study. Indians also get married earlier than people from other Asian or western nations. For example, in 2000, 83.6% of Indian women and 40.1% of Indian men in the 22–24 ages were married; these figures are higher than that in China, Japan, Hong Kong, Korea, Vietnam, or Malaysia and most other western nations (United Nations 2004). This might explain why a higher proportion of both men and women in Indian ads were portrayed in relationship roles.

Other differences may be attributable to religious and other cultural differences between India and the other Asian nations or western nations that have been studied. India is a predominantly Hindu nation (over 80% of the Indian population is Hindu), while most of the other Asian nations are predominantly Muslim, Christian or Buddhist in nature. Hinduism emphasizes that a person has to perform his/her familial (or relationship) roles during the third stage in his/her life (‘grahastya’); in fact, nonperformance of one’s duties as son/daughter, father/mother, etc, invited censure (Chakraborty and Chakraborty 2007). Perhaps this explains the greater use of relationship roles for both men and women in Indian ads.

Secondly, there are other cultural differences between India and other countries (including those in Asia). One theoretical framework used often to analyze cultural differences is Hofstede’s cultural dimensions; India differs significantly from other nations along these dimensions (Hofstede 2001). India scores higher than the Asian average on individualism (44; Asian average: 20) and power-distance (72; Asian average: 60) but lower on long-term orientation (61; Asian average: 95). India’s higher scores on power-distance might, for example, explain the lower ratio of women and the extremely low proportion of female voiceovers in these ads from in ads from other Asian countries. Inequality is more easily accepted in nations with higher power-distance in organizational and familial relationships; this probably explains why the ratio of women (who have lower power in society) as characters is lower in Indian ads than in other Asian nations. This might also explain why the ratio of female voiceovers is lower in Indian ads than in ads from other parts of Asia as voiceovers are considered to represent authority in ads.

India’s score (56) on masculinity—the most dimension most often related to gender stereotyping—is higher than the world average and is very close to the Asian average (53). Nations with lower masculinity are normally expected to place a higher emphasis on relationships, but in this study, the proportion of relationship roles was higher than in less masculine Asian countries such as Korea. Perhaps in the case of India, the impact of the higher masculinity score is counteracted by the strong emphasis on extended or joint family systems in Indian society and the high emphasis the culture places on relationships (Harwood and Roy 1999; Perlow and Weeks 2002). Or, as Odekerken-Schroeder et al. (2001) note, a nation’s masculinity score alone may not explain the level of gender stereotyping in advertisements. Further research is needed to explain the differences between Indian television ads and those from other nations.

Finally, Indian women’s notions of equality and feminism appear to differ from those of other nations—especially western nations. As noted earlier, Indian women see equality in not individual-centered but as family-centered ways and as still inclusive of maintaining valued roles in the home and prizing family above all (Wolf 2009). This might explain the higher prevalence of relationship roles (for both genders) in Indian television ads than in those from other countries. Indian advertisers seem to be focusing on achieving a delicate balance between western and Indian notions of modernity by showing women in relationship roles and in home settings while, at the same time, also depicting them as users of a wide range of products and (when employed) in professional roles. To a great extent, this is a positive trend for as Reddy et al. (2007) note, Indian men and women are more confident in their identities than their parents are and embrace both global trends but, at the same time, do not want to lose their traditions or culture. Marketers have to recognize and leverage these contradictory trends in Indian culture.

Advertisers should be aware of the changing Indian market—especially the changing attitudes of younger Indian women who are exposed more and more to international media and are shedding traditional Indian attitudes toward family and work (Business Week 2005; Kotwal and Sahni 2008; Reddy et al. 2007). For example, Kotwal and Sahni (2008) note that adolescent girls are especially interested in seeing women portrayed as confident, independent, and career-oriented. Thus, marketers may need to revisit their notions of the ‘New Indian Woman’ and portray women more often in career-oriented roles to appeal to the younger female market.

From an international marketer’s point of view, understanding the current gender role portrayals in Indian ads may aid in developing ads that are more suited to the Indian culture. In addition, the differences found between Indian ads and those from other Asian nations emphasize the fact that country-specific ads rather than region-specific ads may be required for success in international marketing. Emphasizing relationships may be more important in the Indian context than in ads for other Asian countries.

As with any study, this study has some limitations. While the channels were chosen based on viewership (the top ones in each category), and both primetime and daytime ads were collected, the entire sample was collected during a 2-week period; a longer data collection period may have yielded a better sample and more generalizable results. Only two Indian language ads were chosen; perhaps ads in other languages may show different patterns. Future researchers should focus on a broader sample with ads collected during a longer period of time.

It should be noted that comparisons across cultures was sometimes made difficult by the differences in measurement. For example, age was measured slightly differently in each study. This study was limited to ads in three languages—English, Hindi and Tamil. It is possible that ads in other languages (and hence other regions of the country) might differ from those in these three languages. In addition, general conclusions are difficult to draw from just one study; more research on Indian ads is needed before generalizations can be made about gender role portrayals in Indian television ads. However, the present study provides an in-depth look at gender role portrayals in one of the world’s largest and fastest growing economies and as such, the findings of this study should be of value to international marketers planning to enter this vast market.