Forecasts underpinning budgetary plans have a well-documented impact on fiscal performance especially when policymakers view the world through rosy lenses. Combined with inertia in the implementation phase, optimism contributes to a possible deficit bias and the accumulation of government debt. This column introduces the new Forecast Tracker of the Secretariat of the European Fiscal Board, which throws light into a less lit corner of EU fiscal surveillance. Covering the period since 2001, it offers systematic and user-friendly access to relevant data and standard accuracy measures for all EU member states.
Forecasts play a pivotal role in fiscal policymaking. They are not simply attempts to predict the future but critical instruments that shape budgetary plans and outcomes. Governments predicate budgets on macroeconomic forecasts, and when economic conditions deviate from original projections there can be reluctance or not enough time to adjust, especially when economic growth falls short. Reluctance may even be justified or desirable to stabilise cyclical downturns and should be followed by a similar reluctance to increase expenditure in the event of revenue windfalls during cyclical booms.
However, research and experience have shown that official forecasts underpinning or guiding government budgets can be biased and overly optimistic. Such a bias, often influenced by political considerations, can result in larger-than-expected deficits (or lower-than-expected surpluses), weaker compliance with fiscal rules, and ultimately affect the sustainability of public finances. Jonung and Larch (2006) were among the first to document a tendency in some EU member states to systematically build budgets on sanguine macroeconomic assumptions and to blame worse-than-expected outcomes on bad luck. Analysing data of a broader host of countries, Frankel (2011) concluded that a propensity towards overoptimism in official macro forecasts can also be observed outside the EU. In addition, Frankel and Schreger (2012) confirm how overly optimistic forecasts in euro area countries help signal compliance with the EU fiscal rules in the planning phase of national fiscal policies while leading to a deferral of fiscal adjustment ex post. Finally, in a more recent study, Larch et al. (2023) corroborate the tendency towards optimism in official budget forecasts in the EU.
Budgetary forecasts and the EU fiscal framework
In 2011, in response to the global economic and financial crisis that shook most advanced economies to their very foundations, the EU strengthened its fiscal framework with the so-called six-pack reform. The reform included specific measures aimed to address optimism in fiscal forecasts. The relevant legal text explicitly stated that “Member States shall ensure that fiscal planning is based on realistic macroeconomic and budgetary forecasts using the most up-to-date information.” Furthermore, the reform required that “[…] macroeconomic and budgetary forecasts for fiscal planning shall be subject to regular, unbiased and comprehensive evaluation based on objective criteria, including ex post evaluation.”
Two years later, the EU decided to strengthen and detail the legal provisions on fiscal forecasts. The so-called two-pack reform of 2013 – applicable to euro area countries – stipulated that national medium-term fiscal plans as well as draft annual budgets had to be grounded in independent macroeconomic forecasts. In particular, macroeconomic forecasts underpinning fiscal plans either had to be produced or endorsed by an independent fiscal council, ensuring a higher level of objectivity and credibility in budgetary projections.
The most recent reform of the EU’s fiscal framework – the one adopted in April 2024 – also touches upon how official macro forecasts underpinning government budget plans are to be produced. However, this time the thrust of the reform is more ambiguous. On the one hand, it extends the obligation to establish independent fiscal councils to all EU member states, i.e. the entities who are meant to instil a sense of realism into the planning process of government budgets. It also requires national governments to follow the conceivably prudent macroeconomic assumptions put forward at the central level by the European Commission, unless they – national governments – have “sound and data-driven economic arguments”. On the other hand, national governments can, but are not obliged to, involve independent councils when defining the path for net expenditure growth over the medium term, the new linchpin of EU fiscal surveillance.
This partial backpedalling on independent expertise can be of significant import in light of the new medium-term focus of EU fiscal planning and surveillance. Under the new rules, fiscal plans and recommendations are not just for one calendar year; they can cover from four to seven years. Naturally, the medium-term orientation amplifies the potential consequences of forecast errors: small errors in successive years can accumulate over time producing a significant impact on public finances. As a result, if initial macroeconomic assumptions underpinning medium-term fiscal plans are too optimistic, a country may be found compliant with the recommended expenditure path while missing initial deficits and debt targets.
Tracking fiscal forecasts in the EU
Against this background, the Secretariat of the European Fiscal Board (EFB) has launched a new tool, the Forecast Tracker. Like its sibling focusing on numerical compliance with EU fiscal rules, 5 the Forecast Tracker aims at enhancing transparency in fiscal surveillance. For reasons that may be obvious, the question of whether or not EU countries manage to stay within the perimeters of the EU fiscal rules attracts most attention in EU fiscal surveillance. Nevertheless, the Forecast Tracker does not enter uncharted territory. There are quite a few studies examining the quality of official fiscal forecasts. Beyond those mentioned in the Introduction, the Directorate-General of Economic and Financial Affairs of the European Commission has, from time to time, performed or commissioned accuracy tests of its own projections (on average, around every four years).
The added value of the new Forecast Tracker is systematic and regular coverage of relevant data combined with conventional accuracy measures and tests – the database will be updated every year. Via a user-friendly application, stakeholders and researchers can select any period or country(-group) of interest. The initial version of the Forecast Tracker is limited to one-year-ahead forecasts of the European Commission of the following key macroeconomic aggregates: nominal GDP, government revenues and expenditure and the budget balance in percent of GDP. In a future evolution, the Forecast Tracker will be extended to the projections underpinning national budgetary plans.
By way of example, Figure 1 offers a very rough idea of the accuracy of the Commission’s one-year-ahead forecast of nominal GDP growth. To clarify conventions, positive (negative) values signal that on average outturns fall short of (exceeded) forecasts. As expected, the average forecast errors vary considerably both across country and time. Starting with the time dimension, forecast errors tend to be particularly negative in years of big crises. This should not come as a surprise, as very few forecasters have the acuity or courage to predict the advent of very large shocks and, once it hits, the exact impact the shock will produce. When major downturns are excluded, average forecast errors on nominal GDP growth are less negative or more positive. Figure 1 also shows a clear procyclical pattern in forecast errors of nominal GDP growth: on average governments tend to be too optimistic when growth is weak or negative, and pessimistic when economies are recovering.
Figure 1 Average forecast errors on nominal GDP growth, year-on-year percentage change (outcome minus forecast) by year and country
Notes: The forecast error is defined as the difference between (i) the outturn in a given year T as recorded in the spring of year T+1 and (ii) the forecast for year T produced in the autumn forecast of year T-1. This formula follows the logic of EU fiscal surveillance where budgetary plans for year T are tyipcally adopted towards the end of the year T-1 and outcomes are checked against EU guidance in spring of year T+1.
Source: European Commission
Turning to the country dimension, there is no immediate pattern when member states are ranked by the size of the average forecast error. Larger member states tend to be clustered on the left where average forecast errors are negative, i.e. they display a degree of optimism. A simple analysis of variance reveals a first interesting result. For countries with an above-average debt-to-GDP ratio the Commission forecasts tend to be a bit more optimistic. The difference compared to countries with below-average debt is significant at conventional levels of confidence for nominal GDP growth and the budget balance in % of GDP.
Beyond the surface
The new tracker also offers the possibility to examine the relationship between fiscal forecasts on the one hand and compliance with the EU fiscal rules on the other. Figure 2 plots the average overall numerical compliance score 7 in 2001-2024 against the average one-year-ahead forecast error of annual nominal GDP growth, and of the budget balance (as a percentage of GDP). The results are consistent with the conventional prior: lower-than-expected GDP growth or worse-than-anticipated budget balances, are associated with a greater likelihood of missing the SGP’s numerical constraints.
Figure 2 Average forecast errors on nominal GDP growth vs average compliance score of EU 27 countries, 2001-2024
Notes: For the definition of forecast errors see notes to Figure 1. The average compliance score is taken from the Compliance Tracker of the European Fiscal Board Secretariat. It measures the average years x rules a country complied with the rules of the Stability and Growth Pact.
Source: European Commission, EFB Secretariat
Beyond looking at simple, two-dimensional correlations, we conducted a series of basic regressions to assess whether forecast accuracy significantly affects compliance with Stability and Growth Pact rules. This approach aims to offer a somewhat more rigorous assessment of the associations identified in the figures above. Our analysis builds on the findings presented in earlier contributions (Larch and Santacroce 2020, Larch et al., 2023).
Figure 3 summarises the results of two types of regression models: (i) logit regressions where numerical compliance is captured by a binary variable: 1 for compliance, 0 for non-compliance; and (ii) linear models where deviations from the numerical constraints of the SGP rules are expressed as a percentage of GDP, where a positive value indicates an overachievement of the numerical constraint of a given rule and a negative value signifies a shortfall. In both types of multivariate inferential models, forecast errors turn out to have a significant effect after controlling for the impact of other relevant variables across the four rules covered in the Compliance Tracker. By way of example, a positive coefficient for forecast errors on nominal GDP growth means the likelihood of complying with a given rule declines (increases) when growth turns out lower (higher) than projected, i.e. when the forecast error is negative (positive). Our results are robust to a number of changes to our specifications.
Figure 3 The estimated effect of forecast errors on numerical compliance
Notes: Logit regression coefficients – a positive value indicates that lower-than-forecast nominal GDP growth weighs on the probabiltiy to comply with the SGP rules. Linear regression coefficients – a positive value means that lower-than-forecast nominal GDP growth is associated with a less positive or more negative deviation from the numerical constraints of the SGP rules.
Source: Own calculations
Conclusions
The new Forecast Tracker of the Secretariat of the European Fiscal Board systematically documents the accuracy of macroeconomic projections used in EU fiscal surveillance, both at the central and, in the future, also the national level. The main aim of the tool is to enhance transparency and support evidence-based fiscal policymaking through annual updates of the database.
A systematic review of forecasts is warranted as persistent optimism can be bad for fiscal outcomes. It can be used to avert difficult decisions in the planning phase while blaming bad luck when worse than expected budgetary outturns are revealed.
Our preliminary findings derived from a first battery of tests confirm the importance of forecast accuracy for fiscal performance in the EU member states: optimistic forecasts affect adherence to the commonly agreed EU rules. The accuracy of forecasts assumes additional importance going forward: Under the reformed EU fiscal framework, national EU guidance and national plans are anchored to forecasts of between four and seven years. By leveraging the Forecast Tracker and its future extension, all interested parties have an additional instrument to make more informed decisions and ultimately contribute to a more effective EU fiscal governance.
Source : VOXeu
