7. There is not yet a proper framework for making decisions about such exemptions in place. In this
paper, various aspects of the nexus between structural reforms and budgetary developments that should
underpin such a framework will be highlighted. In the first part of the paper we look at the following issues:
i) to what extent do budgets cost structural reforms? The United Kingdom, which has a very elaborate fiscal
framework, is used as an example. ii) There may be budgetary cost from not reforming government
programmes, such as disability or early retirement programmes, which should also be taken into account.
iii) Large inventories of structural reforms exist. While they have been used to estimate the effects of
reforms on growth, budgetary implications are seldom quantified. This is obviously difficult to establish.
Two countries, New Zealand and Australia, which have implemented a radical reform agenda are used here
to shed some light on these issues. The second part of the paper uses econometric tools to establish to what
extent structural reform affects public expenditure, in the short as well as the long run. The data allow us to
detect both upfront cost and long-run gains of structural reform, although both seem to be modest in
comparison with other structural determinants. The third part uses simulations with the OECD’s Interlink
model to highlight the implications of reforms that affect technical progress, the participation rate and the
natural rate of unemployment in different settings.
Linking reforms and the budget
The UK budget: a role model?
8. Assessing the budgetary implications of structural reforms in the short, medium and long run is
shrouded with many difficulties. In many cases, governments do provide a costing of changes in tax and
spending plans and often also beyond the current budget.1 Obviously, changes in tax and spending plans are
pursued for many other purposes than purely fiscal ones. UK Budgets (HMT, 2001, 2002, 2003 and 2004),
for instance, provide an assessment of Budget policy decisions over a three-year horizon under the
following headings: meeting the fiscal rules and funding public services; meeting the productivity
challenge; increasing employment opportunity for all; building a fairer society; a modern and fair tax
system; and protecting the environment.
9. Budget 2002 (Table A.I), for instance, lists 54 spending and tax measures and estimates their
budgetary effect. The budgetary impact of 16 measures implemented since Budget 2001 are estimated as
well. The biggest measure was a payroll tax increase and the second biggest an increase in the generosity of
the Child Tax Credit and Working Tax Credit for families with children. All other measures were small,
ranging from beer duty relief for small brewers to simplifying capital gains tax. While the list of measures
is long, only those are included where the impact of the decisions and circumstances can be quantified with
reasonable accuracy. Moreover, spending that is fixed by the spending reviews and embedded in
Departmental Expenditure Limits is not included in the Budget costing of decisions. In 2002, the net fiscal
impact of the identified spending and revenue measures was nearly 1% of GDP, but considerably lower in
most other Budgets. In comparison, the forecasting error for the deficit was equivalent to 1% of GDP for
the year-ahead projection in recent years and nearly 1½ per cent of GDP for the two-year ahead projection
(Table 1). There is no strong hint in the Budget costings that reforms have generated large budgetary costs
in the short term. However, this is so mainly, because the spending underpinning the ongoing reforms to
health, education and infrastructure are built into (hence respect) the Departmental Spending Limits.
1 . According to the 2003 Survey on Budget Practices and Procedures (available at http://ocde.dyndns.org/), 55% of the OECD countries
provide multi-year cost estimates for all new spending items, and another 20% do it for some mandatory spending items. The survey does not
cover revenue changes.
45