Advertising as a Major Source of Human Dissatisfaction: Cross-National Evidence on One Million Europeans

Abstract

Advertising is ubiquitous in modern life. Yet might it be harmful to the happiness of nations? This paper blends longitudinal data on advertising with large-scale surveys on citizens’ well-being. The analysis uses information on approximately 1 million randomly sampled European citizens across 27 nations over 3 decades. We show that increases in national advertising expenditure are followed by significant declines in levels of life satisfaction. This finding is robust to adjustments for a range of potential confounders -- including the personal and economic characteristics of individuals, country fixed-effects, year dummies, and business-cycle influences. Further research remains desirable. Nevertheless, our empirical results are some of the first to be consistent with the hypothesis that, perhaps by fostering unending desires, high levels of advertising may depress societal well-being.

Appendix on Data and Methods

For this paper, data are taken from three different sources: The Eurobarometer Survey, Zenith-Optimedia, and the World Bank. The Eurobarometer survey, which began in 1972, is a set of public opinion surveys conducted on behalf of the European Commission. Each spring and autumn, face-to-face interviews are conducted for a new sample of residents of European Union (EU) Member States (around 1000 per country). The questions that respondents are asked are varied and include items intended to assess life satisfaction, to elicit opinions about the state of politics in Europe, to gain insight into perceptions of political institutions, etc. The data recorded in the Eurobarometer are used by the European Commission to monitor the evolution of public opinion and ultimately to aid in decision making.

For this study, data are gathered from individuals from 27 countries over the years 1980–2011. Specifically, data are available on the following transition European countries: Bulgaria, Czech Republic, Estonia, Croatia, Hungary, Latvia, Lithuania, Poland, Romania, Slovenia, Slovakia, and Turkey, and on the following non-transition countries: Austria, Belgium, Germany, Denmark, Spain, Finland, France, UK, Greece, Ireland, Italy, Netherlands, Norway, Portugal, and Sweden. The survey contains information on individual demographics, such as age, gender, education, marital status, employment status, and household size, as well as life satisfaction indicators. In particular, the survey asks “On the whole, are you very satisfied, fairly satisfied, not very satisfied or not at all satisfied with the life you lead?” Answers to this question are available for every year except 1996.

Annual country total advertising expenditure data are available from Zenith-Optimedia, which is a global media services company. They publish a quarterly report (the “Advertising Expenditure Forecasts”) that covers advertising from a large number of markets around the world. This record contains the total amount spent on advertising in the country historically as well as forecasts for the future. Here historical data are used from 1980 to 2011, as reported in the issue “Advertising Expenditure Forecasts of December 2013.” Further details are available in Austin A, Barnard J, Hutcheon N, Advertising Expenditure Forecasts. Zenith-Optimedia, December 2013.

Macroeconomic indicators are taken from the World Bank. In particular, data are available by country for the years 1980–2011 on GDP, GDP per capita, and the national unemployment rate. These are published in World Development Indicators. Information is combined from all three data sources for the same 27 countries and time periods (1980–2011). The final sample-size for the current study consists of a little over 900,000 observations on randomly sampled European citizens.

The data are used to estimate coefficients from linear regression models, where robust clustered standard errors are computed to account for the fact that the errors may be correlated within countries. Life satisfaction scores are regressed on a variety of control variables as detailed below. Specifically, the main Eq. (10.1) that is estimated is

$$ L{S}_{ijt}=\alpha +\beta AdvEx{p}_{jt}+\Phi Dem{o}_{ijt}+\Gamma Macr{o}_{jt}+{\nu}_j+{\eta}_t+{\epsilon}_{ijt} $$

(10.1)

where i denotes an individual, j a country, and t a year. The variable LS_ijt is reported life satisfaction, AdvExp_jt represents advertising expenditures (measured, in turn, as the lag of natural logarithm of total advertising expenditure and as the sum of three previous lags of natural logarithm of total advertising expenditures), the vector Demo_ijt contains individual demographic characteristics (age, education, gender, etc.), and Macro_jt is a vector of macroeconomic variables that may impact life satisfaction, such as the lag of GDP per capita and the unemployment rate. To control for common country and year attributes, the statistical analysis allows for country (ν_j) and time (η_t) fixed effects. The ϵ_ijt term captures an individual, country, year specific error. A number of different specifications are estimated as robustness checks.