Fiscal devolution and dependency

Abstract

Public spending devolution in practice is widely seen as more appropriate for addressing varied political aspirations within state boundaries than is tax devolution. A drawback is that devolved public spending may be subject to irresistible upward pressure, as illustrated by ‘formula drift’ of the United Kingdom devolved administrations. By crowding out the private sector such public spending can exacerbate the problem it was originally intended to alleviate. When taxpayers do not value increases in government output at least as highly as the private goods and services they must forgo to finance them, then the public sector is too large. This article estimates a three sector Hecksher–Ohlin model of the economy with the greatest relative rise of the public spending ratio in the United Kingdom, Wales. Simulation of the model shows a net gain in employment from a 1% cut in income tax matched by a corresponding reduction in government spending. This result is consistent with the current level of intergovernmental transfers being excessive.

Notes

¹ One part of this condition apparently is satisfied for Wales and Scotland according to the ‘British Identity’ MORI poll for the Economist in November 1999. The Welsh and Scottish have a stronger identification with Wales and Scotland than with Britain, in marked contrast to residents in England.

² Unless all shocks are on the supply side.

³ If increments to what the government was doing were useless in utility terms, they would still record a contribution to output because of the way output is measured.

⁴ PCA, primary current account balance, is the series of output and real exchange rate combinations at which exports equal imports. A given output and real exchange rate determines the demand for imports. If the supply of exports contracts then PCA balance shifts left.

⁵ For Scotland, Midwinter (2002, 108) points out that the Scottish Executive was able to accommodate free personal care for elderly and teachers’ pay increases within Treasury allocated expenditure growth totals, because its share of the UK budget was rising – despite the Barnett ‘squeeze’.

⁶ This is a mark-up on manufacturing (traded) sector wages since the traded sector wage is exogenous.

⁷ For a discussion on the implications of the two models of unemployment determination see, for instance, Blanchflower and Oswald (1994).

⁸ The correspondence with the theoretical sectors is not exact since there is some traded output produced in the nonmanufacturing sector (for example, mining at the beginning of the period and tourism at the end). Historically coal exports have been of major significance for the Welsh economy. On the one hand, relatively little employment and output (11% in the Industrial Production index of 1970) was accounted for by mining even at the beginning of the period. On the other, it was more erratic than manufacturing- employment and output fell earlier (between 1971 and 1974, output declined 40%). So there is an overstatement of buoyancy of export sector in the early years.

⁹ It is well-known that the OLS estimation of equations with lagged dependent variables, as some equations in our model are, renders the DW statistics unreliable.

¹⁰ Explaining productivity in UK manufacturing as a whole has proved problematic, partly because apparent productivity slowdown in the 1970s stemmed from incorrect measurement of output and from structural change (Cameron, 2000).

¹¹ The figure reported is for the year 2001. For example, for the manufacturing sector the cost is £418/week*0.01*52 weeks per employee. For both sectors combined the cost per job is £31 180.