Green Is Less Greedy : Competition, Corruption, and Productivity in Green Public Procurement in Bulgaria

Although green public procurement has been established as a desirable policy goal across the globe, especially in the European Union, its scope and impacts remain severely understudied. This paper provides insights into the prevalence and structure of green public procurement in Bulgaria, which is a sustainability laggard within the European Union and hence a least likely champion of green public procurement. The paper also estimates the impacts of green procurement on traditional procurement and economic outcomes: competition, corruption risks, and overall productivity. Using novel data and more comprehensive methods than previous studies, the analysis finds that green public procurement amounted to about 10 to 20 percent of total public procurement spending in Bulgaria in 2011–19. Most descriptors and requirements of green public procurement are found in titles, technical requirements, and product descriptions. Green criteria in award criteria texts, which are mainly used for flagging green public procurement in the literature, have been marginal in comparison. Green public procurement is estimated to improve competition for government contracts among firms, for example by increasing the prevalence of market entrants by 3 to 7 percentage points. Green public procurement contracts are also less prone to corruption risks. For example, they are 0.6 to 1.5 percentage points less likely to receive a single bidder. Finally, green public procurement enhances the efficiency of resource allocation in the economy by helping to channel public resources to more productive firms, for example to those that have 14 percent higher labor productivity. This effect is at least in part explained by the positive interaction between green public procurement and the lower risk of corruption. The findings strengthen the case for pursuing green public procurement goals as they offer synergies with traditional public procurement goals.