There is a vast amount of literature about gender portrayals in television advertisements. This type of gender study first emerged in the United States in the 1970s during the second wave of the feminist movement (Dominick & Rauch, 1972; McArthur & Resko, 1975). Two decades later, these were followed by some English-language studies on Southeast and East Asian countries including Japan (Arima, 2003; Bresnahan, Inoue, Liu, & Nishida, 2001; Milner & Collins, 2000), China (Cheng, 1997; Siu & Au, 1997), Taiwan (Bresnahan et al., 2001), South Korea (Kim & Lowry, 2005; Prieler, 2012), Malaysia (Bresnahan et al., 2001; Tan, Ling, & Theng, 2002; Wee, Choong, & Tambyah, 1995), Singapore (Siu & Au, 1997; Tan et al., 2002; Wee et al., 1995), and Hong Kong and Indonesia (Furnham, Mak, & Tanidjojo, 2000). Nearly all of these studies concluded that television advertisements portray gender, and especially women, in stereotypical ways (Furnham & Paltzer, 2010).

Despite the large quantity of studies on gender portrayals in television advertisements, and on the aging of many societies around the world, to date there have only been two publications of quantitative content analyses that dealt explicitly with gender representations of older people (Prieler, Kohlbacher, Hagiwara, & Arima, 2011; Stern & Mastro, 2004). In this study we define older people, in accordance with Japanese advertising agency conventions, as people over the age of 50 years (Dentsu Senior Project, 2007; Hakuhodo Elder Business Suishinshitsu, 2006). This study was the first to analyze gender portrayals of older people in Japanese television advertising over an extended period using data from 1997 and 2007–a time period in which the number of older people in Japanese society increased particularly strongly. This study provided an opportunity to investigate whether gender representations of older people in Japan have changed over one decade, and if advertisements reflect any social changes, including the rising importance of older people in Japanese society. In addition, this study is also able to show if there are any gender differences in the representation of gender in general and gender representations of older people in particular. Finally, possible effects of gender representations are discussed in terms of social cognitive theory and cultivation theory.

Since there is nearly no quantitative content analysis that explicitly deals with gender portrayals of older people, we refer in the following literature review to research on gender and older people, especially those studies published after 2000 on Asian societies. Deriving hypotheses from this diverse body of literature posed several problems, such as cultural and national differences; the variety of sampling methods (recording times, geographical area, channels) and units of analysis (TV ads vs. TV show characters); sample size; and wide variations in the definition of the term older people, which can range anywhere from 45 years to over 65 years of age. While we cannot superimpose our results on these previous findings, they nevertheless functioned as a basis for developing hypotheses and variables for our own research.

The results of this study are based on chi-square analyses that were executed on a sample of unduplicated television advertisements that included older people. Because we were interested in identifying the overall significant differences between 1997 and 2007 for each category, and which subcategories contributed to this difference, we broke the results down even further using the adjusted standardized residuals (ASRs) for post-hoc tests. In turn, we will try to answer each hypothesis while simultaneously addressing RQ1 (regarding the changes between the years 1997 and 2007) through conducting chi-square tests for each gender.

Between 1997 and 2007, there was a significant increase in the number of advertisements depicting older males (from 93 to 153), as well as the number of ads showing older females, (from 52 to 65). Despite an increase in depictions of both genders over the decade study period, older females were much less represented than older males in both years (1997: χ² = 11.593, df = 1, p = .001; 2007: χ² = 35.523, df = 1, p < .001). This proved hypothesis 1 for both years, which predicted the depiction of more older males than older females in the advertisements.

In terms of the setting (see Table 1), gender differences emerged for both years (χ² = 12.173, df = 4, p = .016 for 1997; χ² = 9.875, df = 4, p = .043 for 2007). Breaking the results further down, using the adjusted standardized residuals (ASRs) for post-hoc tests, showed that significantly more males than females appeared in the outside setting in 1997 (26.9% vs. 9.6%, ASR = ±2.5) and in the workplace setting in 2007 (18.3% vs. 4.6%, ASR = ±2.6). However, while there was a higher percentage of females than males in the home setting, these results were not statistically significant, as indicated by the adjusted standardized residuals (32.7% vs. 19.4% for 1997, ASR = ±1.8; 21.5% vs. 12.4%, ASR = ±1.7 for 2008). Thus, hypothesis 2a, which stated that more older females than males will be shown in the home setting, was rejected. Hypothesis 2b, which stated that more older males than older females will be shown in the workplace setting, was supported for 2007, but not for 1997. Comparing the 1997 and 2007 results found a significant increase in depictions for males (χ² = 11.299, df = 4, p = .023), but not for females (χ² = 7.398, df = 4, p = .116). As compared to 1997 data, in 2007 more males were depicted in the workplace (18.3% vs. 6.5%, ASR = ±2.6) and less in the outside setting(16.3% vs. 26.9%, ASR = ±2.0).

Analyzing the social interaction of older people showed that in 2007 no overall gender differences emerged (χ² = 2.246, df = 4, p = .717), although in 1997 (χ² = 9.315, df = 4, p = .046) a gender difference was found as more females than males being depicted alone (30.8% vs. 16.1%, ASR = ±2.1). Thus, hypotheses 3, which stated that more older males than older females would be shown alone, was not supported by the data. Changes emerged between 1997 and 2007 for males (χ² = 24.895, df = 4, p < .001), but not for females (χ² = 2.716, df = 4, p= .449). More males were shown alone in 2007 (45.1% vs. 16.1%, ASR = ±4.6) and less with other adults (38.6% vs. 63.4%, ASR = ±3.8).

The image of older people showed no gender differences in 1997 (χ² = 1.185, df = 2, p = .553), while gender differences did emerge in 2007 (χ² = 11.843, df = 2, p = .003). More specifically, more males then females were depicted in a favorable way in 2007 (50.3% vs. 33.8%, ASR = ±2.2) and more females than males were shown in a neutral way (60.0% vs. 35.3%; ASR = ±3.4). Thus, hypothesis 4 stating that older males will be depicted in favorable ways more often than older females was supported in 2007, but not in 1997. Comparing the two years, no differences for favorable depictions of females emerged (χ² = 1.895, df = 2, p = .388), but more favorable depictions of males were found in 2007 (χ² = 6.346, df = 2, p = .042) and there were a smaller percentage of males in the neutral category in 2007 than in 1997 (35.3% vs. 51.6%, ASR = ±2.5).

In terms of the product category, gender differences were found for 1997 (χ² = 27.655, df = 6, p <. 001) as well as for 2007 (χ² = 43.276, df = 6, p < .001). More specifically, in 1997 more older females than older males were shown in advertisements for the product categories of cosmetics/ toiletries (23.1% vs. 6.5%, ASR = ±2.9) and services/leisure (26.9% vs. 4.3%, ASR = ±4.0). Conversely, for other product categories, more older males than older females were depicted (38.7% vs. 19.2%, ASR = ±2.4). In 2007, more older males were depicted in advertisements for food/beverage (27.5% vs. 10.8%, ARS = ±2.7) and more older females for cosmetics/ toiletries (35.4% vs. 3.9%, ASR = ±6.3). Thus, hypothesis 6, which stated that there will be more older females than males in advertising for the cosmetics/toiletries product category, was supported. When 1997 and 2007 results for the product category service/leisure were compared, no differences emerged for older males (χ² = 7.117, df = 6, p = .310), although differences were found for older females (χ² = 12.930, df = 6, p = .035) with less older females being portrayed in service/leisure advertising in 2007 than in 1997 (7.7% vs. 26.9%, ARS = ±2.8).

Overall, gender representations of older people in Japanese television advertisements were rather stereotypical in both 1997 and 2007. For example, more older females were depicted in advertising for the cosmetics/toiletries products in both years, and more older males than older females were shown in television advertisements. However, while the representations of older females did not change much between the two years, the portrayal of older males became increasingly more dominant and more positive. While males were already dominating in 1997, even a higher percentage of older males than older females were depicted in 2007, more males than females were shown in the workplace in 2007, and males were being represented in more favorable ways than females in 2007.

In both years, depiction of older males in advertisements clearly outnumbered the depiction of older females. The percentage of older females depicted even decreased between 1997 and 2007 (from 35.9% to 29.8%). This is a surprising result, given that older women outnumber and outlive men in Japan. However, this finding confirms that negative attitudes toward older women still exist in Japan (Formanek, 2008). This discriminatory under- representation of older females was already decried in the 1970s in the United States as the “double standard of aging” (Sontag, 1972, p. 29) and over the past 40 years not much seems to have changed in Japan. This mis- and under-representation of older females has real consequences in that it can shape society’s attitudes towards older women (Gerbner, 1998; Gerbner et al., 1980). For example, people might learn through such representations that older women are not as valuable as older males and as younger females (Bandura, 2009).

Older people were not only misrepresented in terms of gender and numbers, but also in terms of the settings in which they were shown. Older females were shown more often than older males at home (32.7% vs. 19.4% for 1997; 21.5% vs. 12.4% for 2007, though not statistically significant). Older males were shown more often in outside settings and in the workplace in 2007, while surprisingly more older females than older males were shown in the workplace in 1997 (though not statistically significant). Thus, the stereotypical gender depictions of older people in terms of the setting increased between 1997 and 2007. In short, setting in Japanese television advertisements depicted the two genders in a stereotypical division of roles with more males at the workplace and more females at home. Japanese television advertisements mirror the traditional stereotype that women should stay at home and take care of the family even though this traditional stereotype does not reflect current attitudinal changes within the Japanese population. Our findings on setting are consistent with most previous research on gender representations (Furnham & Paltzer, 2010).

Results for social interaction differences were rather mixed and not clear-cut. While more older females than older males were shown alone in 1997, the situation changed in 2007 when both males and females were predominantly shown alone and not portrayed in the company of other older adults as in 1997. At first sight, these findings might imply that representations of older people has become more aligned with the stereotype of older people as being lonely. However, this result could also be interpreted as a positive sign – namely that older people are finally being regarded as important enough to be shown alone and do not need to be depicted with other age groups. There might be various reasons for these results that showed an increase in the portrayal of older people being alone. As previously shown (Prieler, Kohlbacher, Hagiwara, & Arima, 2010), more than 60% of older people in Japanese television advertisements are celebrities. Thus, the predominance of older people being shown alone might be also connected with the fact that the majority of older people depicted were celebrities.

The overall image of older people also changed between 1997 and 2007. While no significant gender differences were identified in 1997, in 2007 significantly more older males than older females were depicted in a favorable way (50.3% vs. 33.8%). In short, over the 10 year period there was an increasing discrepancy between the favorable representations of males and females. The fact that the image of older people became even more stereotypic and differentiated by gender from 1997 to 2007 might come as a surprise, but this result is hardly unexpected since Japan is still a highly patriarchal society where men generally play more important roles than females (Sugimoto, 2003) and are regarded as having a higher status than females in society.

Last but not least, the gender representations of older people in terms of the product category were most stereotypical as exemplified by the cosmetics/ toiletries product categories using more older females than males in both 1997 and 2007. Social cognitive theory suggests that we can learn from such associations (Bandura, 2001). In other words, the strong association between females and cosmetics/toiletries emphasizes the importance that society assigns to female beauty and contributes to their sexualization (Luyt, 2011). Previous research, in such diverse places as Japan, Malaysia, Taiwan, India, Indonesia, and Singapore, has shown that such findings are not only true for older females, but for females in general (Bresnahan et al., 2001; Das, 2011; Furnham et al., 2000; Tan et al., 2002). This association was not only prevalent in Asia, but also the most consistent finding around the world (Furnham & Paltzer, 2010).

Overall, such stereotypical gender representations of older people, which increased between 1997 and 2007, are problematic since advertisements and other media can help to define gender and age roles, something that has been pointed out by the social cognitive theory and the cultivation theory (Bandura, 2009; Gerbner, 1998). Empirical research supports these theories. Advertisements have been shown to influence gender role attitudes (Garst & Bodenhausen, 1997; MacKay & Covell, 1997) and media influences the way in which older people perceive themselves (Donlon et al., 2005; Mares & Cantor, 1992) as well as how older people are perceived by younger people (Gerbner et al., 1980). Thus, the representations of older people as well as gender are important social issues with real consequences.

Gender differences of older people in Japanese television advertisements were examined for 1997 and 2007. Surprisingly, gender differences increased between the two years with more stereotypical gender representations in 2007. While the occurrence of older males in advertisements outnumbered the number of older females in 1997, this difference became even more pronounced in 2007. In addition, more older males than older females were depicted in the workplace in 2007 with the older males being represented in a more favorable way. For both years, the cosmetics/toiletries product category was predominantly advertised using females. Overall, such representations might have negative effects on the audience by perpetuating gender stereotypes.

This study has been one of the first to shed light on the increasingly important question of how older people are represented by gender in television advertisements. At a time when many societies around the world, including in Asian countries, are dealing with an increasingly aging population, this research area should no longer be neglected. Our study has answered some questions, but much still remains to be done. As is the case for every content analysis, we can only claim representativeness for the two periods of time analyzed (1997 and 2007). As a follow-up study, we suggest future investigation of audience reaction to these representations of older people in order to gauge the possible social effects. In conclusion, the representation of older people in advertising is a research topic ripe for further examination, and such research stands to significantly benefit gender and women’s studies as well.