In this study, carried out for the Federal Ministry of Finance, the Ifo Institute reviewed the methods used for estimating potential output and the use of the resulting output gap for the calculation of structural (i.e. cyclically adjusted) budget balances. This breakdown is important since the actual budget balance reflects the influence of both temporary effects due to cyclical deviations of output from trend and permanent factors reflecting the discretionary decisions of the government. Estimation of the structural balance involves three steps. In the first step potential output and the output gap are estimated, in the second step the cyclical component of expenditures and revenues are quantified using appropriate defined elasticities, and in the third step the structural component is estimated by subtraction of the cyclical component from the actual deficit.
In the study, the different methods proposed by national and international agencies (German Council of Economic Advisers, OECD, EU, IMF) and in the literature are first analyzed from a methodological viewpoint. In principle, two different approaches can be distinguished. The first is the production function approach adopted by the OECD and the IMF. The second approach relies on statistical time series methods to decompose an observed series in trend and cycle. The dominating method in this second group is the Hodrick-Prescott (HP) filter, preferred by the European Commission. We especially analyzed the endpoint problem and the revision of the output gap estimates when new data become available. We found for several European countries that the HP filter needs typically 3 to 4 years before the estimator converges to a stable figure. Compared with the first estimate for a given year, the revision of the output gap amounts to up to 3 percentage points. Thus, there is substantial uncertainty about the current cyclical state of the economy.
We then estimated an Unobserved Components Model for all EU countries and compared the results with those obtained from the HP filter and the production function approach (OECD, IMF). In a qualitative sense, all methods draw a similar picture of the cycle, but the numerical results differ in some years by more then 2 percentage points. There are several cases where one or two methods indicate a positive output gap whereas the other methods indicate a negative gap. The conclusion is that there is high uncertainty and risk concerning the diagnosis of the precise value of the output gap. Estimates of output gaps may be a useful tool in the hand of experts but are not suitable as a basis for strict fiscal rules.
The cyclical component of the fiscal deficit is usually calculated by multiplying the output gap by an appropriately defined net expenditure elasticity. This elasticity can be estimated on an aggregate level or be constructed by a combination of disaggregated revenue- and expenditure elasticities. The study shows that the different approaches used by the various institutions reach approximately the same results. In order to get reliable elasticities, it may be important to adjust the raw data for changes in tax legislation.
RSS