14 September 2005, 01:27
Social Marketing in Perm: Children's Rights
Theodore P. Gerber and Sarah E. Mendelson
This paper addresses the results of the June 2004 regional survey assessing the impact of the children's rights campaign on residents of Perm. While there are other measures of the impact, we address here the quantitative results from the resurvey in Perm and the control group Sverdlovsk. The activists in Perm ran two different campaigns: one that focused on children's rights with the phrase "ya chelovek" and one that focused on foster care and was primarily aimed at government officials ("deti dolzhny"). We report on both campaigns here. The campaigns appear quite successful in terms of how many residents they reached. We find, however, limited impact in terms of shifting attitudes. We suspect the messages were not clear enough, and we do not find they provoked the desired reaction. Again, as in the Ryazan case study, we find no evidence of indirect effects of the campaign on attitudes.
Exposure and Reaction to the "Ya Chelovek..." Campaign (Table 1)
Finding: About a third of Perm residents were exposed to the campaign and over 80% of them understood what it was about.
Based on our data, 28% of the population of Perm city saw the "Ya chelovek" billboards, 6% saw the calendars, and 15% saw the TV programs on at least one occasion. Thus, the billboards appear to have reached the broadest sample in the city. Combining the three media of distribution, one third of the city residents (34%) had some exposure to the campaign. The campaign also reached a fairly large percentage of the population outside the capital in Perm oblast, primarily via television: 14% saw at least one TV program, 19% overall encountered the campaign at least once in one form or another. Even allowing for some "false positives" (erroneous statements of respondents that they saw the campaign), these results indicate the "Ya chelovek" campaign reached a considerable proportion of the population.
It appears that it may even have reached some in Sverdlovsk oblast. Although we suspect that most of the respondents from Sverdlovsk who claim to have seen the billboards (7%) or calendars (6%) are false positives, it may be that the television programs were broadcast in parts of Sverdlovsk and did reach about 10% of the population there. Almost the same proportions of Sverdlovsk respondents who say they saw the materials correctly identify them as dealing with children rights (41%) as identify them as part of an election campaign (38%), yet the former figure is large enough to lend some credence to the possibility that many of those Sverdlovsk respondents did indeed see the television programs.
Correct interpretation of the campaign materials was very high in Perm city: 84% identified them with rights of children (Panel E). In non-capital Perm, the figure was 62%. These numbers are impressive: they suggest the theme of the campaign was clear, especially when we consider that many of those in non-capital Perm who did not correctly identify the theme could well be "false positives" who never actually saw the materials.
Judging by the reaction of the survey respondents in Perm city, the campaign got its message across very effectively: 68% strongly agree and 21% agree with the statement that when they see the "Ya chelovek" materials they want more to be done to protect children's rights. The numbers for non-capital Perm and for Sverdlovsk are equally impressive. Altogether, the "Ya chelovek" campaign must be judged a success in terms of the number of people it reached and the clarity of the theme and message.
Who saw "Ya chelovek..." and who reacted as hoped (Table 2)
Finding: Younger residents understood the campaign better than older residents.
Our logistic regressions revealed that younger and better-educated Perm city residents were more likely to see the "Ya chelovek" campaign materials, while gender had no significant effect. If the campaign's objective was to reach those under 30, it accomplished this: the coefficient of .846 implies their odds of seeing the materials were 2.3 times higher than the odds for those over 30. Among those who saw the campaign, those under 30 were also significantly more likely to correctly identify the theme as dealing with children's rights. No other variable significantly affected this outcome, and none of the variables (cohort, gender, or education) had a significant effect on the reaction to the materials.
Exposure and Reaction to the "Deti Dolzhny..." Campaign (Table 3)
Finding: The smaller "deti dolzhny" campaign reached a smaller share of the population than the "ya chelovek" campaign although we find it still reached 20%.
Our results suggest that 15% in Perm city saw the ads on the street at least once and 13% saw the ads on public transportation. Still, 20% overall report seeing some form of the campaign, an impressive number given the size of the campaign. We also note here that only 7% in Perm oblast outside of the capital and only 8% in Sverdlovsk report seeing the campaign materials. The fact that these figures are substantially lower than the figures for "Ya chelovek" increases our confidence that many of the positive responses regarding "Ya chelovek" outside Perm city reflect actual exposures to the campaign rather than false positives.
The "Deti dolzhny" campaign was also quite effective in terms of communicating its theme: in Perm city, 83% of those who saw the campaign materials correctly identified the theme, compared with 76% in non-capital Perm and 59% in Sverdlovsk (Panel D). The campaign provoked the desired reaction for two-thirds of those who saw the materials in Perm city (Panel E), 63% in non-capital Perm, and 65% in Sverdlovsk. Based on these numbers, the vast majority of those who saw the materials understood what they were about and a sizable majority responded in the manner expected by the campaign planners.
The children's day activities were somewhat less successful. In Perm city, nearly two-thirds of the respondents did not hear about the children's day activities (Panel F), and only 4% actually participated. The numbers who say they knew of the activities were higher outside of Perm city, but the level of participation similarly low.
Who saw "Deti dolzhny" and who reacted as hoped? (Table 4)
Finding: Again, younger residents appear to have understood the campaign better.
As in the case of the "Ya chelovek" campaign, we found that those under 30 were more likely to say they saw the "Deti dolzhny" materials. Also, the college educated had greater exposure to the campaign. Thus, both campaigns were especially successful at reaching younger and college educated residents of the city. Correct identification of the campaign theme and reaction to the campaign did not vary at all by cohort, gender, or education. In Perm city, women were more likely to have heard about the children's day events (Panel D). They were also much more likely to attend them: in fact, not a single man participated in the children's day events (thus, for technical reasons the "woman" dummy variable could not be entered in the logistic regression model.) Clearly, the organizers of the children's day events need to do more to make the activities appeal to men, unless they are content to involve women only.
Dynamics of questions from the Perm group specifically relating to children's issues (Table 5)
Finding: The campaigns had mixed impact on general attitudes relating to children's issues and failed to increase support for government assistance to children.
The campaigns sought to increase public concerns about children's rights in general and support for foster care as a way to deal with the problem of orphans and homeless children. Aggregate patterns in survey responses to the questions measuring these attitudes offer some indication of whether the campaigns attained these goals. The campaign can be judged a longer-term success to the extent that we see some movement in the aggregate on these questions, particularly in Perm city, where the campaign exposure was greatest.
In fact, from February 2003 to July 2004, there is no significant change in the proportion of Perm city residents who say that adolescents should have the same rights as adults (Panel A). There is an increase in the proportion endorsing this position in non-capital Perm, though, from 51% to 81%. But we see no movement in either of the two Perm samples with regard to assessing whether the rights and worth of adolescents are respected in Russia today (Panel B). Altogether the impact of the campaign on general orientations toward children's rights is minimal.
The impact on views of specific violations of children's rights is more evident. In Perm city, opposition to corporal punishment for children increased from 83% to 90% (Panel C). Opposition to taking away a toy was up slightly - from 87% to 90% (Panel D). Rejection of the practice of reading a child's letter without permission grew from 88% to 96% (Panel E). Even though disapproval was already quite high in Perm city, the campaign may have pushed it even higher. Similar, if somewhat weaker, movement is evident in Perm oblast, where disapproval of corporal punishment went from 83% to 88% and disapproval of taking away a gift from 90% to 94%, while opposition to reading a letter remained stable at 93%-94%. The fact there was virtually no movement - or even some movement away from disapproval - on these questions in Sverdlovsk bolsters our confidence that the movement observed in Perm may be related to the campaign.
In the 2003 survey activists asked us to test the slogan "children get all the best!" and how it applied to the Soviet-era and/or post-Soviet reality. While the campaign did not actively use this slogan, we had to resurvey in 2004 on this question. In hindsight, we might have been concerned that by drawing attention to current problems involving children, the campaign could actually have increased nostalgia for the Soviet-era and thus would have been orthogonal to the goals of the human rights activists in Perm. But from February 2003 to July 2004, the proportion saying that the slogan applied to Soviet times either completely or somewhat remained stable at 61%-62% in Perm city, though there was a noteworthy shift from the "completely" to the "somewhat" category. Endorsement of this position fell in non-capital Perm from 76% to 59%. At the very least, these numbers reassure us that the campaign did not have the unintended consequence of making Soviet-era reality seem better by drawing attention to problems involving children in contemporary Russian society.
In fact, the campaign does not appear to have dimmed people's view of children's situation in contemporary Russia. In Perm city, 81% felt that the slogan applied only a little bit or not at all in 2003, down to 77% in 2004 (Panel G). No change is evident in non-capital with respect to this question. If the campaign aimed to increase the level of general concern about children's issues in contemporary Russia, it seems not to have much impact based on this measure.
The data show an increase in the proportion of Perm city residents who say that the family (from 68% to 80%) and the school (46% to 57%) are among the most important spheres influencing the development of children (Panel H). It is not immediately evident that the campaign meant to accomplish this result. But it could be that thinking about children's issues in response to the campaign led some citizens of Perm to place greater emphasis on these crucial sites. The number for the school also increased noticeably in non-capital Perm.
When we turn to the question regarding which vulnerable group the government should help most (as the top priority), the results suggest the campaign did a poor job communicating its message. Specifically, while the campaign sought to increase the proportion advocating government aide for children, that number actually decreased from 27% to 19% in Perm city (Panel I). Support for priority assistance to the elderly also decreased - instead, support for helping the poor and "hard to say" both grew. In non-capital Perm, support for prioritizing assistance to children grew slightly - from 24% to 28%. Still, we do not see much evidence that the campaign produced greater public demand for the government to take action to help children.
The "deti dolzhny" campaign explicitly sought to increase support for adoption and foster care as solutions to the problems of orphans and homeless children. Unfortunately, there is no evidence of any impact on public opinion in Perm city: 24% saw aiding foster parents as the best solution in 2003, 25% in 2004 (Panel I). It is noteworthy that this measure increased from 16% to 24% in non-capital Perm, but we doubt this can be attributed to the campaign, since only 7% of respondents outside the capital report encountering any of the "deti dolzhny" materials.
Finally, we do not see any notable change in the amount of time that respondents with young children spend interacting with their kids (Panel K).
Overall, the "ya chelovek" and "deti dolzhny" campaigns seem not to have had much impact on how the population perceives problems relating to children's rights. They may have increased disapproval about specific practices that violate children's rights - corporal punishment, rescinding a gift as punishment, and snooping on private materials. But they did not increase general support for the rights of adolescents or the level of concern about the situation of children or demand for government intervention on behalf of children.
Assessing campaign impact at the individual level (Table 6)
Finding: Our individual-level analysis confirmed our impression from the analysis of aggregate trends that the larger impact of the Perm campaigns was limited.
If the Perm campaigns did not have much impact at the aggregate level, they nonetheless may have had some influence on those who saw the materials and correctly understood their themes as relating to children. For those observers interested in quantitative representation of the campaign's impact at the individual level, we provide below a detailed discussion of a series of multivariate analyses of the individual-level determinants of answers to the questions analyzed in the previous section.
We first examined the determinants of the probability of choosing "the lack of a future for our children" as one of the five or six most serious problems facing the country (see below). Our initial model showed that concern about children is higher among women and lower among both the oldest and youngest cohorts. Otherwise, our model reveals no significant change in Perm city or anywhere else between 2003 and 2004. Our complete model revealed no individual-level effects of exposure to the campaign; however, this may result from the collinearity among the main and interaction effects in Model 2. When we limit the measures of exposure to the interaction term, women*ya chelovek, we observe a weakly significant positive effect. This effect implies that women who saw and understood the theme of the "ya chelovek" campaign materials have 55% (e.441 = 1.55) higher odds of saying that lack of a future for children is one of the country's top problems than women who did not see (or did not correctly identify the theme of) the "ya chelovek" campaign. In turn, women who did not see the campaign have 48% higher odds of expressing concern over children's future than men - whether or not the men saw the campaign materials.
At the individual level, the data suggest the campaign increased agreement that adolescents in principle have the same rights as adults (Panel B). The negative, statistically significant coefficient implies that exposure to the campaign raised the probabilities associated with higher levels of agreement that children have rights. This is the proper interpretation because lower values correspond to response categories indicating agreement - for this analysis and all others, we coded "hard to say" responses as neutral, i.e., as falling between "somewhat agree" and "somewhat disagree." It is difficult to translate the coefficients from ordinal logit models into intuitively meaningful terms, so we do not attempt to do so here. Suffice it say that there is a statistically significant association between comprehending the theme of the "ya chelovek" campaign and supporting the notion that adolescents have the same rights as adults, controlling for demographic and locational factors that may also affect views on that proposition.
As for those other factors, we note that those over 50 are more likely to agree that adolescents have rights, as are residents in Perm city in 2003 (the main effects of Perm city and Perm other pertain to 2003 - the effects in 2004 can be obtained by adding the main effects to the interaction effects involving the location and 2004). Overall, the level of agreement that adolescents have rights decline (based on the positive, significant main effect of 2004), but it increased in non-capital Perm, and it increased among those who saw the campaign materials. Thus, the multivariate, individual-level models suggest the picture is more complex than the aggregate analyses in the previous section imply. The campaign may have been more effective than is evident from aggregate trends: for those who came into contact with it, it helped offset a general tendency away from agreement that children have rights.
At the individual - as at the aggregate - level the campaign does not appear to have affected views on whether children's rights and worth are respected in Russia today (Panel C). Turning to specific practices that violate children's rights - measures where we saw evidence of campaign impact in the aggregate analyses - we find that exposure to the "deti dolzhny" campaign is associated with decreased support for corporal punishment of children (Panel D). This is consistent with the aggregate Finding, but it is a bit surprising that exposure to "deti dolzhny" had an impact, while exposure to "ya chelovek" did not. Moreover, neither campaign had a statistically significant effect on views toward either of the other practices at the individual level (Panels E and F). All in all, these results are disappointing: because of the aggregate trends that suggested some campaign impact on these measures, we would have expected statistically significant effects at the individual level.
Based on our multivariate models, those who saw and understood the "ya chelovek" campaign are more likely to disagree that the "All the best for children!" slogan applied to Soviet reality (Panel G). Perhaps the use of "rights" language in reference to children effectively dissuaded those who saw the campaign from taking an overly rosy view of the situation of children in the Soviet era. Neither campaign influenced agreement that the statement applies to current reality (Panel H).
We also find no evidence that, at the individual level, the "ya chelovek" or "deti dolzhny" campaigns influenced people's views on whether support for foster parents is the best approach to the problem of orphans and homeless youths (Panel I) or whether the government should first and foremost try to help children as opposed to other vulnerable groups in Russian society (Panel J).
We see some effects in the expected direction on some variables, but in most cases the effects of exposure to the campaigns are not statistically significant. Although the Perm activists successfully reached fairly large numbers of people with their billboards, calendars, television programs, and other campaign materials, the evidence that these materials shaped people's views regarding children's issues is weak at best.
Dynamics of views on HR NGOs, problems facing the country, and human rights (Tables 7-9)
Finding: We see no indirect effect of these campaigns on other attitudes.
The aggregate trends suggest the campaign did not do much to increase support for NGOs that defend human rights in general (Table 7, Panel A), nor did it bolster support for Memorial in particular (Panel B). In Perm city, positive views on HR NGOs in general declined from 66% to 60%, while familiarity with Memorial and assessment of its activities did not change significantly. While positive views toward HR NGOs increased in non-capital Perm from 59% to 75%, at the same time knowledge of Memorial declined from 26% to 14%. These results can only be viewed as disappointing: the campaign did not yield benefits for Memorial or for HR NGOs more generally.
Nor did the Perm campaigns increase the general level of concern about the lack of future for children (Table 8): the proportion choosing this concern as one of the five or six most pressing problems in the country did not change significantly from 2003 to 2004 in Perm city, Perm oblast, or Sverdlovsk oblast. We do observe in Perm city growing concerns over poverty, economic crisis, and access to health care - but no change regarding the salience of these problems outside of Perm city. At any rate, it is not possible that the campaign provoked concerns over these issues. Also, in Perm city worries over drug addiction and HIV/AIDS declined.
Finally, when we look at trends in support for the nine specific human rights, we see minimal and inconsistent changes over time (Table 9). In sum, there is no evidence that the Perm campaigns increased broad support for human rights.
In sum, the Perm campaigns appear to have reached a satisfactory number of people, most of whom understood that the materials dealt with problems involving children and most of whom say that when they saw the materials they wanted children's problems to be addressed or wanted to learn more about them. In these respects, the campaigns were successful: they reached a broad audience, and most got the basic idea.
However, in terms of broader impact, the message appears to have had little results. There is very little evidence - at either the aggregate or individual level - that the campaigns affected the level of public concern about issues involving children. There were some exceptions, but overall we conclude that the messages of the campaign were not effective at producing the shifts the activists had hoped for in public opinion.
 We pooled the data from the 2003 and 2004 surveys, and began with models incorporating demographic variables (for differences by cohort, gender, education, and rural vs. urban residence), dummy variables for Perm city and non-capital Perm (contrasted with Sverdlovsk), the year 2004 (contrasted with 2003), and interactions between region and 2004 to capture possible differences in the shifts between 2003 and 2004 in Perm city and non-capital Perm. We then added dummy variables denoting exposure to and correct identification of the themes of, respectively, the "ya chelovek" and "deti dolzhny" campaigns. Because we suspected that the campaigns may have had a stronger impact on women, we also tested interactions involving campaign exposure and gender. For those dependent variables that have Likert-like scales as response categories, we use ordinal logistic regression; in other cases, where the dependent variable is dichotomous, we use simple logit models.