Annual nominal expenditure ceilings are set three years in advance as part of the budget process, and are
considered to be binding. The ceilings apply to central government primary expenditure, including
transfers and grants to local governments, plus expenditures by the old-age pension system outside the
central government budget. Each year, as part of a rolling budget framework, an additional ceiling is
applied to expenditures three years out.^16 The ceilings for years t+1 and t+2 could in principle be altered,
but this has not happened since the system was adopted in 1997 (except for technical adjustments). The
ceilings are set with a margin over projected expenditures to allow for some policy flexibility and, more
importantly, for increases in cyclical spending during an economic downturn. An attempt by parliament
to change a proposed budget has to be presented in the form of a complete package that respects the
previously determined expenditure frames and ceilings. This requirement has strengthened the hand of the
minister of finance in the budget process and has made it more difficult for the budget to be defeated or
amended in parliament.
Nominal expenditure ceilings have been an effective means of achieving the surplus target in Sweden. In
fact, the ceilings together with a prolonged economic upswing, where revenue collections continuously
exceeded projections, produced surpluses that exceeded 2% of GDP between 1999 and 2001. As a result
of the expenditure ceilings, fiscal headroom produced by this boom was saved or used for tax cuts rather
than for expenditure increases. However, the margins for cyclical fluctuations have been fully used during
economic upturns even though they were intended to be only a safety cushion during unexpected
downturns. As a result, the ceilings came under pressure following the 2002-03 downturn, forcing the
government to scale back some expenditure commitments. The habit of using all headroom under the
ceiling for expenditure increases and using the ceiling more as an expenditure target is worrisome and has
contributed to a general government surplus lower than 2% of GDP since 2002, but still the ceiling has
been important in reducing the expenditure ratio for the central government in the late 1990s and after
that keeping it at a stable level.
Apart from the tendency to use up the margins for expenditure, Sweden’s fiscal framework has two
potential weak spots. First, expenditure restraint has been less evident at the local level, where most
government consumption takes place, than at the central government level. Second, the government has
resorted to the limited use of tax expenditures to introduce new policies without breaching the ceiling or
requiring balancing measures.
(^16) Between 1997 and 2001 the ceiling for t+3 was set by parliament in the spring (March-May). Since 2002 it is instead
proposed in the Budget Bill and decided in the autumn (September-November). In autumn 2004 no ceiling was set for
- Instead, the government planned to propose ceilings for both 2007 and 2008 in the Budget Bill for 2006 (in
autumn 2005).