Public procurement: award mechanisms and implementation process

This thesis investigates the choice of award mechanism and the process of implementing the award of contracts in public procurement. Public procurement is an activity conducted by the public sector to purchase goods and services. Government spending on public procurement accounts for 10% to 25% of GDP in each country (World Bank Group, 2017). Because of its large sharing in government expenditure, public procurement can shed light on important practical policy issues, including those as investigated here: the choice of award mechanism (i.e. the process for selecting contractors); the possibility of costly delays in awarding contracts; and the concern that corruption may lead to inefficient outcomes. This thesis uses public procurement data for the UK and other EU countries plus Iceland and Norway. These countries adopt the same benchmark award mechanisms (the EU benchmark award mechanisms). The EU benchmark award mechanisms are implementations of the award mechanisms defined by the World Trade Organisation (WTO) Agreement on Government Procurement (GPA), which influences the award mechanism arrangement in 88 countries. There are four EU benchmark award mechanisms: the open procedure, the restricted procedure, the negotiated procedure and the competitive dialogue. Few empirical studies have investigated these EU benchmark award mechanisms. This thesis contains three independent and interrelated studies. The first study examines the choice of award mechanism in the UK, using the logit model. The results show that UK public buyers choose award mechanisms that are consistent with the theoretical suggestions. When a contract is complex, a UK public buyer is likely to employ an award mechanism that allows for greater discretion of selection. It also provides evidence that public buyer's experience is an important factor in award mechanism choice. Also based on UK data, the second study compares the decision speed of awarding a contract, using the logit model and survival analysis. It uncovers that delay in contract award is a serious issue for public procurement, as almost half of the contracts in our sample experienced delays in contract awards. The empirical results show that the negotiated procedure (which contains negotiations) is likely to be associated with a more rapid decision speed and a lower probability of delay than the restricted procedure (which does not contain negotiations). Therefore, this study casts doubt on the general expectation of practitioners that negotiation causes delays. The third study explores the relationship between discretion, corruption and competition in public procurement. Based on the revenue equivalence theorem and extensive form game, it proposes a game theory model showing that discretion fosters corruption, which in turn depresses the number of bidders and softens price competition. This is a mechanism that the procurement literature agrees but few formal models exist. The OLS estimates show a negative correlation between corruption and the number of bidders and suggest that procuring agents may disguise the impact of corruption by inflating the estimated contract value.