(McPherson 2003), and most accounts of structural inefficiencies at these companies
have been largely anecdotal.
This paper analyses the operating and financial performance of privatised NOCs,
based on a dataset of 60 share-issue privatisations (‘SIPs’) by 28 firms (from 20
countries) in the period 1977 to 2004. For each firm, a total of 22 different metrics is
calculated in order to comprehensively capture firm performance and efficiency.
Privatisation here is understood to be the initial sale of (part of) the government equity
interest to private investors, where the government has been the controlling
shareholder prior to that sale.2 This definition hence includes both partial and full
privatisations via the equity markets, but excludes privatisation sales to other industry
buyers. For the sample of initial SIPs, we first employ a univariate testing
methodology to compare the pre- and post privatisation performance levels of
privatised firms. Secondly, in order to move beyond this simple comparison, we also
investigate the time pattern of changes through a multivariate panel data regression
analysis. Although the focus on initial SIPs is very common in comparable
longitudinal studies3, privatisation is usually undertaken via multiple offerings with
the government being unlikely to transfer control in the very first offering. We
therefore extend – in a third step – the time horizon of analysis to include any possible
follow-on offerings of the respective oil and gas companies.
The remainder of this paper is structured as follows: Section II briefly reviews the
existing literature, Section III describes the dataset of global share-issue privatisations
in the oil and gas sector; Section IV analyses the performance impact associated with
initial SIPs; Section V focuses on follow-on SIPs; Section VI discusses some
potential concerns as to the study design; Section VII concludes.
2 3 ‘State’ and ‘government’ ownership are used interchangeably in this paper. In fact most previous studies restrict themselves to the analysis of initial SIPs. We are not aware of other studies which consider all privatisation offers over time for a select group of companies.