Measuring the effects of advertising on green industry sales: a generalized propensity score approach

Abstract

ABSTRACTThis article estimates the effects of advertising expenditures on annual gross sales of green industry firms using a quasi-experimental framework. In order to account for potential selection bias, a generalized propensity score and a dose-response function are used to estimate advertising treatment effects. The method used allows us to investigate the relationship between the dose (advertising expenditures) and the response (firm sales). We use data from the National Green Industry Surveys of 2009 and 2014 to conduct the analysis. To further investigate potential heterogeneous advertising effects of the size of the firms, we separate the sample into small firms and large firms, according to their annual gross sales. The results indicate that the magnitude and shape of the response function depend on the size of the firm. For small firms, increasing advertising spending yields to higher sales within a range of advertising spending. Beyond this range, advertising spending increases do not impact sales any more. Thus, small firms’ management should carefully monitor advertising input. For large firms, on the other hand, the current evidence does not support a positive relationship between advertising spending and sales since the marginal treatment effect is insignificant almost over the entire range of adverting spending.