3. Empirical evidence
Rail
Some of the problems following the liberalisation in the 1990s of rail tracks in the United
Kingdom have been blamed on massive under-investment. A more detailed analysis of the
underlying figures suggests however no systematic under-investment in comparison with the
previous system at the sectoral level. Pollitt (2000) finds that in constant 1995 prices between
1981 and 1995 investment increased significantly in British Rail (as well as in airports and in the
water industries, while no significant changes took place in telecommunications, electricity and
gas industries). Investment in rail transportation did not drop but rather increased in 1999
following privatisation in 1997. Furthermore, the investment rate was high in historical terms
after privatisation (compared to the 1970s and 1980s) and the introduction of incentive regulation
(Clark, Elsby and Love, 2001).14 Overall, incentive regulation did not appear to cause a major
disruption in sectoral investments of the regulated sectors. The overall investment figures in the
British railway sector until 1999 mask, however, a steady increase in investment in the railway
network and a fall in investment in rolling stock (Affuso and Newbery, 2000).15
Telecommunications
There are a number of studies that show a strong positive relationship between investment and
incentive price regulation in the US telecommunication industry. For instance, Ai and Sappington
(2002) note that past research has shown that incentive regulation was more powerful than
traditional rate-of-return regulation during the late 1980s and early 1990s for the deployment of
14. Measuring investment at constant prices or as a percentage of GDP may not appropriately
capture the quality dimension of investment. For instance, Clark, Elsby and Love (2001) show
that while being stable over the 1980s, total investment in roads in the United Kingdom picked
up between 1990 and 1994 and then sharply declined from 1994 to 1999, in part because road
maintenance became more efficient over time. Kinnunen (2006) argues that the decline in real
investment in Finnish electricity distribution after liberalisation is not necessarily an indication
of lower volumes of physical investment. The same amount of network can be built at lower cost
if new technology or the implementation of it becomes less costly.
15. Affuso and Newbery (2000) point out two possible reasons for a decline in rolling stock
investment. First, contracts awarded to the train operation companies shorter than the lifetime of
the rolling stock discourages investment. Second, in accordance with real option theory, train
operators may postpone investment in the face of uncertainty with regard to future demand.
Affuso and Newbery (2000) analyse the determinants of discretionary investment (investment
other than committed to the regulator) of 25 train operators between 1997 and 1999. Their
estimation results indicate that shorter contracts promote discretionary investment and that
higher demand uncertainty and higher profits are positively correlated with discretionary
investment.
10