excess burden of taxation and so minimise the distortionary costs of taxation. Angeletos (2002) and
Bnera and Nicolini (2004) show how this outcome can be achieved even in the absence of a complete
set of contingent securities, by exploiting variations in the yield curve across different maturities
of risk free securities. Critical to this approach is the idea that debt management can exploit
a negative covariance between adverse fiscal shocks and bond prices. When governments receive
news of adverse future fiscal trends a fall in bond prices helps maintain the intertemporal budget
constraint - the market value of government debt equals the net present value of future primary
surpluses with minimal need for tax rates to change. We term the ability of governments to insulate
tax rates (and the excess burden of taxation) in this manner fiscal insurance. The aim of the paper
is to evaluate the potential of a wide range of possible indicators of fiscal insurance, construct
estimates of fiscal insurance using OECD data and then see whether there is any relationship
between realized degrees of fiscal insurance and the composition of government debt.
Our notion of fiscal insurance is also relevant to the case where the aim of debt management
is not to stabilize the excess burden of taxation but rather to stabilize or target the level of debt,
possibly so as to maximize the probability of governments achieving stated fiscal rules (see Giavazzi
and Missale (2004), Borenztein and Mauro (2004), Goldfajn (1998), Lloyd Ellis and Zhu (2001) for
a similar motivation to debt management). We therefore use both optimal tax smooth and debt
stabilization as possible motivations for why governments will be interested in the fiscal insurance
potential of debt management and the various tests we propose.
A distinctive feature of our analysis is an empirical evaluation of observed practices rather than
a normative analysis leading to a recommended optimal portfolio. A substantial literature exists
using the optimal taxation paradigm to draw normative recommendations regarding the optimal
portfolio structure for government debt (see, inter alia, Bohn (1990), Barro (1997), Missale (1999),
Angeletos (2002), Lustig, Sleet and Yeltekin (2005), Nosbusch (2008)). While offering many insights
this approach provides few implications for analysing existing portfolios. Further, as Faraglia,
Marcet and Scott (2007) show the predictions of these analyses are often dramatically different from
observed debt portfolio structures. In the absence of performance indicators for debt management
there is no way of knowing whether these sharp differences in portfolio structure actually lead to