SRI: Retrospective verification of medium- to long-term GDP forecasts

The Sompo Institute Plus (SRI), the think tank of Sompo Group, has published a report that analyzes how accurate government economic growth forecasts have been in Japan, particularly focusing on projections made in 2015 for the period through 2023. It examines why these forecasts often missed the mark, finding that:

1. Actual economic growth fell significantly short of projections (achieving only 1/3 of the optimistic "Economic Revitalization Case")
2. Total Factor Productivity (TFP) growth was much lower than expected, dragging down overall growth
3. Labor participation rates actually exceeded even the optimistic projections
4. Government forecasts were revised in 2018, but still remained too optimistic

The report concludes by examining the latest forecasts from January 2025, which include a new middle-ground "Growth Transition Case," but suggests these may still be overly optimistic about productivity improvements and capital investment.

Introduction and Background

In today's complex economic environment, characterized by global climate change, domestic population decline, and concerns about fiscal and social security sustainability, medium to long-term economic forecasts have become increasingly important. The Cabinet Office of Japan regularly produces "Medium to Long-term Economic and Fiscal Projections" twice yearly, providing economic outlooks for approximately the next decade. These projections serve as reference materials for the Council on Economic and Fiscal Policy.

The Sompo Institute Plus report conducts a retrospective verification of these forecasts, specifically examining economic growth projections made in February 2015 for the period through fiscal year 2023. This verification is valuable because it allows to assess the accuracy of past forecasts, identify patterns in deviations, and determine how promptly corrections were made when forecasts diverged from reality.

Verification Methodology

The analysis focuses on three key questions:

1. To what extent did past economic growth forecasts deviate from subsequently determined actual values?
2. Which specific component items primarily contributed to these deviations?
3. When were course corrections made in response to identified deviations?

The report examines projections from February 2015, which contained two distinct scenarios:

• The "Economic Revitalization Case" - an optimistic high-growth scenario assuming successful economic policy implementation
• The "Baseline Case" - a more conservative scenario based on the potential growth rate at that time

For tracking corrections, the analysis follows forecast revisions from January 2016 through January 2019, noting that in 2018, the "Economic Revitalization Case" was renamed the "Growth Realization Case."

Major Findings on Forecast Accuracy

Overall Growth Performance

When comparing actual real GDP growth rates from fiscal 2016 through 2023 with the 2015 projections, the analysis reveals significant underperformance. The cumulative growth over this period achieved only about one-third of what was projected in the optimistic "Economic Revitalization Case" and fell substantially short of even the more conservative "Baseline Case."

While the COVID-19 pandemic naturally caused significant temporary economic disruption that no forecast could have anticipated, the analysis suggests that even without the pandemic, Japan's economy would likely have underperformed both scenario projections. By fiscal 2023, when post-pandemic economic normalization had largely run its course, cumulative growth rates remained significantly below forecasted levels.

Component Analysis

When breaking down the sources of economic growth using growth accounting principles, the report identifies contrasting patterns between productivity and labor inputs:

TFP (Total Factor Productivity) Contribution:

• TFP growth, which represents technological progress and efficiency gains, severely underperformed from the beginning
• The actual TFP contribution fell significantly below even the "Baseline Case" projections
• The "Economic Revitalization Case" had assumed TFP would rise to approximately 2.2% (the average from February 1983 to October 1993) by the early 2020s, which proved wildly optimistic

Labor and Capital Contribution:

• Despite the overall economic underperformance, labor and capital inputs actually exceeded the "Baseline Case" projections
• The labor participation rate rose dramatically, surpassing even the optimistic "Economic Revitalization Case" assumptions
• This strong labor market performance was driven by increased participation among women and the elderly, primarily in non-regular employment, combined with declining unemployment rates

This analysis reveals that the primary cause of the forecast's excessive optimism was unrealistic expectations about productivity growth. Despite stronger-than-expected labor inputs, persistently weak productivity growth dragged down overall economic performance.

Timeliness of Forecast Revisions

The report also evaluates how promptly government projections were adjusted when deviations became apparent. Signs of TFP underperformance and labor participation overperformance were evident from early on.

TFP Forecast Revisions:

• Until 2017, the "Economic Revitalization Case" continued to assume TFP would rise to around 2.2% by the early 2020s
• Significant downward revision occurred in 2018 when the scenario was renamed the "Growth Realization Case"
• The "Baseline Case" assumptions were slightly revised downward in 2019

Labor and Capital Forecast Revisions:

• Similar patterns emerged, with meaningful revisions occurring in 2018 to reflect stronger labor participation
• Both the high-growth scenario and the baseline scenario were adjusted during this period

The report suggests that these revisions were implemented in a relatively timely manner, considering several constraints:

1. Major annual changes to long-term projections could create market confusion
2. A certain observation period is necessary to confirm whether changes represent established trends
3. Political considerations often make fundamental forecast alterations difficult without changes in administration policy

Despite these revisions, comparisons between the 2019 forecasts and actual results for fiscal 2020-2023 reveal that an optimistic bias remained in the projections, though the magnitude of deviation had decreased compared to earlier forecasts.

Latest Forecast Evaluation

The report concludes by examining the most recent projections published in January 2025, assessing whether lessons from past forecast errors have been incorporated. Several positive developments are noted:

1. The former "Growth Realization Case" has been renamed the "High Growth Realization Case"
2. The former "Baseline Case" has been renamed the "Past Projection Case"
3. A new intermediate scenario called the "Growth Transition Case" has been introduced

The addition of this middle-ground "Growth Transition Case" represents a positive step, providing a more realistic outlook that acknowledges challenges while maintaining reasonable optimism. This scenario assumes stable growth exceeding 1% in real terms as Japan exits deflation.

However, the analysis identifies potential areas where excessive optimism may persist:

TFP Assumptions:

• Even in the "Growth Transition Case," relatively high productivity growth is assumed (reaching an annual rate of around 1.1%, the average over the past 40 years)
• Current TFP growth remains around 0.5%, raising questions about the feasibility of this assumption

Labor and Capital Assumptions:

• The "Past Projection Case" shows continued gradual deceleration, leading to negative growth in the 2030s
• The "High Growth Realization Case" and "Growth Transition Case" assume a temporary recovery toward the end of the 2020s
• These optimistic scenarios rely heavily on expanded capital investment under a growth-oriented economy driven by wage increases

The report notes that the breakdown of potential growth in these forecasts shows that while labor input contribution remains flat from fiscal 2029 to 2034, the contribution of capital input is projected to expand significantly. The report questions whether this capital investment optimism is warranted, suggesting that upcoming data will relatively quickly reveal whether this assumption is realistic.

Implications and Significance

This analysis carries several important implications for policymakers, economists, and market participants:

Persistent Optimism Bias: The consistent pattern of overly optimistic economic projections, particularly regarding productivity growth, suggests a systematic bias in governmental forecasting. This pattern persisted despite periodic revisions to align with emerging realities.

Productivity Challenge: The significant underperformance of TFP growth highlights Japan's persistent productivity challenge. Despite various policy initiatives aimed at enhancing efficiency and innovation, actual productivity improvements have remained elusive.

Labor Market Resilience: The stronger-than-expected performance in labor participation demonstrates unexpected resilience in Japan's labor market, even amidst population aging and decline. This suggests potential underestimation of adaptability in labor market dynamics.

Forecast Credibility: Persistent overestimation of growth potential raises questions about the credibility of official economic projections, which serve as foundations for fiscal planning, monetary policy, and private sector investment decisions.

Policy Effectiveness: The gap between projected and actual economic performance may reflect limitations in the effectiveness of economic policies designed to boost growth and productivity.

Conclusion

The retrospective verification of Japan's medium to long-term economic forecasts reveals a pattern of persistent optimism, particularly regarding productivity growth expectations. While labor market participation exceeded projections, weak productivity growth consistently dragged down overall economic performance.

The introduction of a more balanced "Growth Transition Case" in recent forecasts represents a positive development, acknowledging past forecast errors and providing a more realistic middle ground. However, even these revised projections may contain elements of excessive optimism regarding future productivity improvements and capital investment.

The analysis underscores the importance of continual verification of economic forecasts against actual outcomes. Such verification enables both greater accuracy in future projections and more effective policy formulation based on realistic assessments of economic potential. Given Japan's ongoing challenges with an aging population, high government debt, and productivity limitations, accurate forecasting becomes increasingly crucial for addressing long-term fiscal sustainability and economic vitality.

As the report concludes, future economic data will relatively quickly reveal whether the latest forecasts have sufficiently incorporated lessons from past projection errors, particularly regarding assumptions about productivity growth and capital investment. Continued vigilance and willingness to adjust projections based on emerging trends remain essential for effective economic policy planning.