Little is known about the possible impact of an influenza pandemic on a nation's economy. We applied the UK macroeconomic model 'COMPACT' to epidemiological data on previous UK influenza pandemics, and extrapolated a sensitivity analysis to cover more extreme disease scenarios. Analysis suggests that the economic impact of a repeat of the 1957 or 1968 pandemics, allowing for school closures, would be short-lived, constituting a loss of 3.35 and 0.58% of GDP in the first pandemic quarter and year, respectively. A more severe scenario (with more than 1% of the population dying) could yield impacts of 21 and 4.5%, respectively. The economic shockwave would be gravest when absenteeism (through school closures) increases beyond a few weeks, creating policy repercussions for influenza pandemic planning as the most severe economic impact is due to policies to contain the pandemic rather than the pandemic itself.Accounting for changes in consumption patterns made in an attempt to avoid infection worsens the potential impact. Our mild disease scenario then shows first quarter/first year reductions in GDP of 9.5/2.5%, compared with our severe scenario reductions of 29.5/6%. These results clearly indicate the significance of behavioural change over disease parameters.

No abstract

The Fiscal Stability Pact for EMU implies that constraints on ®scal policy facilitate in¯ation control. In this paper we identify two stable policy regimes. When monetary policy seeks to raise real interest rates in response to excess in¯ation, a self-stabilising ®scal policy is required to ensure model stability. A ®scal policy which does not, by itself, ensure ®scal solvency constrains monetary policy to be relatively`passive'. However, in simulations we conclude that the central bank does not need to seek, on this account, the degree of debt stabilisation that appears to be implied by the ®scal stability pact.

Fiscal Sustainability in a New Keynesian ModelRecent work on optimal monetary and fiscal policy in New Keynesian models suggests that it is optimal to allow steady-state debt to follow a random walk. In this paper we consider the nature of the time inconsistency involved in such a policy and its implication for discretionary policymaking. We show that governments are tempted, given inflationary expectations, to utilize their monetary and fiscal instruments in the initial period to change the ultimate debt burden they need to service. We demonstrate that this temptation is only eliminated if following shocks, the new steady-state debt is equal to the original (efficient) debt level even though there is no explicit debt target in the government's objective function. Analytically and in a series of numerical simulations we show which instrument is used to stabilize the debt depends crucially on the degree of nominal inertia and the size of the debt stock. We also show that the welfare consequences of introducing debt are negligible for precommitment policies, but can be significant for discretionary policy. Finally, we assess the credibility of commitment policy by considering a quasi-commitment policy, which allows for different probabilities of reneging on past promises. JEL codes: E62, E63Keywords: New Keynesian model, government debt, monetary policy, fiscal policy, credibility.IN STICKY-PRICE NEW KEYNESIAN models where social welfare is derived from consumers' utility, Benigno and Woodford (2003) and SchmittGrohe and Uribe (2004) show that optimal steady-state debt follows a random walk when policymakers can commit to a time-inconsistent monetary and fiscal policy. In this paper we analyze the nature of the time inconsistency involved and show thatWe would like to thank the editor, Paul Evans, three anonymous referees, Javier Andres, Florin Bilbiie, Tatiana Kirsanova, Luisa Lambertini, Ioana Moldovan, Raffaele Rossi, Martin Uribe, and participants at seminars at the Universities of Oxford and Valencia and the Bank of Spain for very helpful comments. We would also like to thank Tatiana Kirsanova and Christoph Himmels for providing us with computer code to facilitate the solution of the model under quasi-commitment. All errors remain our own. We are also grateful to the ESRC, grant nos. RES-156-25-003 and RES-062-23-1436, for financial assistance.