Oil nationalisations as protracted affairs: evidence from Venezuela

• Date: 01 March 2016
• DOI: http://doi.org/10.1111/opec.12061
• Author: Jorge E Portillo
• Published date: 01 March 2016

Oil nationalisations as protracted affairs:

evidence from Venezuela

Jorge E Portillo

Deputy Manager, Department of Economic Research, Central Bank of Venezuela, Caracas 1010,

Venezuela. Email: jportill@bcv.org.ve

Abstract

Empirical studies tend to treat oil nationalisations as basically discrete, sudden events, an approach

that in some cases might be misleading and adds little to disentangle causality. On a contrary vein,

we take a closer look at the 1970s nationalisation of Venezuela’s oil industry and find evidence that

this was a rather protracted affair,expanding over a 17-year period. Weargue that although the trans-

fer of all assets to government hands was not completed until 1975, oil companies fullyanticipated

such move as early as 1958 and started to operate in a way compatible with a much shorter time

horizon. Specifically,we use an estimate of the resource rent (defined as output price minus marginal

extraction cost) to show a major shift in oil firms’ implicit time horizon after 1958. Our results help

to explain the rather weak causality link found in the data between oil nationalisations and price

fluctuations.

1. Introduction

Starting in the late-1990s, there has been a fresh wave of forced nationalisations of major

oil assets around the world, which in turn has rekindled the academic interest on the

subject.1A common feature in all these studies is the use of a rather static analytical frame-

work, where nationalisations are basically one-sided, discrete events: Every period, the

host government weighs its options so that when the gains of nationalising outweigh the

costs, then oil assets are seized; otherwise, it is business as usual.As a case in point, Guriev

et al. (2011) argue that, in line with the relational contracts literature, there is an inefficient

equilibrium where, in response to a positive oil price shock, governmentswill nationalise

oil assets rather than impose higher taxes.They use panel data on nationalisations in the oil

industry around the world during 1960–2006 to estimate conditional logit and probit

models with country fixed effects and find that governments are more likelyto nationalise

when oil prices are high (relative to the long-term trend) and when ‘political institutions

are weak’, in the sense that the government cannot crediblycommit itself not to national-

ise. On a contrary vein, Kennedy and Tiede (2011) find little qualitative and quantitative

evidence to support a causal connection between high oil prices and nationalisations and

©2016 Organization of the Petroleum Exporting Countries. Published by John Wiley & Sons Ltd, 9600 Garsington

Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA.

50

instead they suggest that ‘politics drive nationalizations’. Specifically, they use the same

dataset as Guriev et al. (2011) to estimate a logit model where the eventof nationalisation

is a function of various proxies for political conditions, finding empirical support for three

main political drivers of oil nationalisations: first, the relative importance of the oil sector

in the overall economy, in the sense that the pursuit of a nationalistic agenda and broader

development goals increase the political rational for control of strategic sectors; second,

the role of policy diffusion, as political entrepreneurs reduce over time the information

problem of would-be nationalisers; and third, the duration of the political regime, in the

sense that younger political regimes are more likely to nationalise as a way to dilute any

political alliance between the international oil companies and the old political elites.

In this context, we take a closer look at the 1970s nationalisation of Venezuela’s oil

industry and find evidence that this was a rather protracted affair,expanding over a 17-year

period, so that reducing such complex dynamics into a single date (when the handover of

assets finally took place) is misleading and adds little to disentangle causality. In that

sense, our research lends support to Kennedy and Tiede’s methodological stand which

favours causal process observation. Specifically, we argue that although the transfer of all

oil assets to government hands was not completed until 1975, the international oil compa-

nies anticipated such move as earlyas 1958 (when an executive order brought to a halt the

leasing of new fields for the exploration) and started to operate in a waycompatible with a

much shorter time horizon.

Our main hypothesis is that international oil companies perceived the 1958 shift in

leasing policies not as a mere threat of nationalisation by the Venezuelan government,

but rather as an ‘early notification’ and adjusted their time horizon accordingly.A similar

claim is advanced by Cakir-Melek (2014) who uses data on the Venezuelan oil industry

to calibrate a dynamic model of a representative foreign firm and simulate the response

to 1961 ‘news’ that a nationalisation was expected to materialise by 1970. The model

predicts that foreign firms would increase extraction effort and reduce exploratory

efforts in anticipation of nationalisation, although no indication is given as to how firms

came about such ‘news’ and the adjustment predicted bythe model is slower than what is

actually observed in the data. In any case, this is only a calibration exercise and does not

provide econometric evidence about the shift in the firm’s time horizon, which is the

focus of the present study.

Westart by pointing out that if an exhaustible resource is to be optimally extracted, the

value of the remaining stock of resource has to be zero at the end of the last year of opera-

tion. This can be achievedby either depleting all the stock available or by driving down to

zero the resource rent (defined as output price minus marginal extraction cost) as the firm

approaches the time horizon. In fact, in the presence of a declining output price, it might be

necessary to increase production over time in order to raise the marginal extraction cost

fast enough to drive the resource rent towardzero, which is actually what happened during

©2016 Organization of the Petroleum Exporting Countries OPEC Energy Review March 2016

Oil nationalisations as protracted affairs 51