Leaders ’ ‘ Green ’ Posts
taken into consideration to identify which leaders focused most on environmental issues on Facebook and whether such a higher focus applied to communicative choices made by these political leaders .
We opted for bivariate analyses for three main reasons : ( i ) it represents the best analytical tool for answering our research questions ; ( ii ) in this article we are not interested in determining the factors explaining a higher focus on environmental issues on Facebook , but we are interested in hypothesis testing tasks of relationship between two single variables ; and ( iii ) the limited number of our cases ( 89 ) does not advice to run multivariate analyses , such as multivariate regressions . A possible choice could have been to conduct a QCA with our dependent variable as outcome and our six independent variables as six conditions in a fsQCA . However , since we did not find any theoretical reasoning for hypothesizing the presence of equifinality or conjunctural causation , we decided to not follow this possible analytical path .
Findings
9,676 out of 25,151 posts ( 38.5 % of the total ) included a reference to a policy issue . When leaders published a post about a policy issue they focused particularly on social ( 31.0 %) and economic ( 25.6 %) policies . On the Facebook pages of the leaders taken into consideration , environmental posts appeared only in 2.2 % of the cases , while this percentage value would rise to 5.8 % when considering the posts about policy issues only . This means that if a citizen hypothetically uses Facebook only to follow the main political leaders of contemporary democracies , he / she will find information about environmental issues only sporadically , and this is true even when looking at the subset of cases in which leaders talk about policy issues .
Table 5 lists the percentages of posts about environmental issues per leader . The mean ( 5.1 %) and the median ( 3.8 %) values confirm the trend pointed out by Table 4 , while the standard deviation ( 6.4 ) expresses the presence of a moderate level of variability : most of the leaders revolve around the mean and median values ; 10 out of 89 leaders reach double digits values ; 11 leaders published no posts at all about environmental issues ; there is only an outlier , that is the Brazilian leader Marina Silva ( z = 7.5 ), whose posts shared about environmental issues constituted 52.9 % of the total number of posts she had shared on policy issues .
T tests analyses and bivariate correlations highlighted in Table 6 draw a complex scenario . First , gender and the effects of the crisis never reach statistically significant values . It does not make any difference on the propensity of publishing environmental posts whether the leader is a male or a female , or whether he / she acts in a country where the economic crisis of 2007 had a strong effect over economic well-being levels measured in terms of GDP . Second , only party affiliation and age are related to an environmental salience if considering all the leaders included into the data set : leftwing and younger leaders are the ones that publish the most posts about environmental issues . Third , Table 6 points out a clear divide when splitting up the analysis between Western and Latin-American leaders . While the independent variables show statistically significant values when conducting the analysis on Western leaders only , no independent variable affects the green agenda of the main political leaders when looking exclusively inside the Latin-American context . Regarding Western leaders ,
80