FEU Special Topic: Forecast evolution and future trends

Introduction

This special topic briefly examines how the Treasury's economic forecasts have evolved over the past 5 years - a period of much greater volatility than in the period prior to the COVID-19 pandemic. A key reason behind the changes to recent forecasts has been a reassessment of future labour productivity prospects[1]. Additional data available since the Budget Update indicates further downside risk to the productivity outlook.

Forecasts evolve for a number of reasons

Economic forecasts evolve over time for a variety of reasons. These include unexpected shocks such as pandemics, adverse weather events, global financial shocks and geopolitical events. Changes to policies, including the response to large shocks such as the pandemic, can also affect the economic outlook. Large swings in commodity prices can be the source of either positive or negative shocks depending on the commodity and direction of change.

As all forecasts are subject to inherent uncertainty, emerging economic data will shed light on whether near-term economic forecasts remain appropriate or whether risks have shifted to the upside or downside. It also provides information as to whether judgements about future trends remain valid. In addition, prior economic data are also subject to revision which can change our understanding of the recent past.

Changes in the economic outlook flow through to changes in fiscal forecasts, the impacts of which can be compounded or offset by changes in expenditure and revenue policies.

Nominal GDP forecasts drive tax forecasts

Forecasts of nominal GDP are a key driver of forecasts for tax revenue. Nominal GDP depends both on how many goods and services are produced in an economy as well as the prices of these goods and services.

COVID-19 resulted in large downward revisions…

Figures 1 and 2 show the evolution of Treasury's forecasts for real and nominal GDP over the past 5 years, including the forecast immediately prior to COVID-19 arriving in New Zealand (HYEFU 2019) and each subsequent Budget forecast. The variation in forecasts over this period was larger than was the case over the prior five years. The onset of COVID-19 saw large downward revisions to Treasury's forecasts in BEFU 2020.[2]

Figure 1 - Forecasts of real GDP

Sources: Stats NZ, Treasury

Figure 2 - Forecasts of nominal GDP

Sources: Stats NZ, Treasury

…although forecasts were subsequently revised higher as the economy showed greater resilience

By 2021 it became clear that while COVID's economic impacts were large, they were not as severe as initially anticipated. Real GDP was revised up but not to pre-COVID levels. However, higher inflation meant that nominal GDP forecasts were back near pre-COVID levels.

Forecasts were revised higher in 2022 with demand proving resilient and bouncing back quickly from periods of heightened restrictions. Inflation accelerated rapidly, owing to both strong demand and disrupted supply, which boosted forecasts of nominal GDP.

High interest rates were required to control inflation, slowing growth

From late 2022 forecasts of nominal GDP have generally weakened. A period of slower growth was expected to be necessary to reduce inflationary pressure and global growth had also slowed. In addition, recent data on GDP per hour worked, or labour productivity, influenced our assessment of longer-term growth prospects.

Earlier burst of productivity growth not sustained…

Over 2021 estimates of labour productivity had surged. They remained at levels well in excess of the pre-COVID-19 period throughout 2022. This strength was projected to continue into the forecast period. However, by the end of 2023 and into 2024 it was becoming increasingly clear that the COVID-19 productivity surge was not being sustained. The Treasury's HYEFU 2023 forecasts revised down productivity over the forecast period, followed by a larger downward revision in the latest Budget Update (BEFU 2024).

Current downturn appears deeper with recovery later than presented in the Budget Update…

Recent data and sentiment suggest that the economic downturn has been deeper, and the recovery may begin later, than forecast in May's Budget Update. Demand remains weak, with retail sales volumes falling slightly in the September quarter and indicators of manufacturing and service activity remaining at contractionary levels. Overall, evidence suggests little or no growth in the economy over recent months.

A feature of our talks with businesses in October was that current conditions remained tough, although there was more optimism for the future. This was consistent with the NZIER's most recent Quarterly Survey of Business Opinion that firms were more pessimistic about their current trading conditions than they have been since 2009 (excluding the initial COVID-19 period). Recent improvements in the ANZ's business outlook also point to greater optimism about the future. This suggests that the anticipated recovery is coming, albeit a little later than previously expected.

…with current estimates signalling further downside to trend productivity

With growth in hours worked falling only slightly in the September quarter, it is likely that labour productivity will have more in common with the pre-COVID slowdown rather than the initial productivity surge during COVID-19 (Figure 3). As we accumulate more such observations our view of trend labour productivity tends to gradually shift downwards and consequently, we see downside risk emerging to our Budget Update (BEFU 2024) assessment for trend productivity.

Figure 3 - Labour productivity trends

Sources: Stats NZ, Treasury

Such assessments are not unique to New Zealand. The Reserve Bank of Australia (RBA)'s November statement highlighted how their forecasts of productivity have been evolving over recent years (Figure 4). Figure 4 illustrates the considerable uncertainty associated with forecasts of productivity. In particular, the RBA has been revising down labour productivity forecasts since 2021, after initially placing too much weight on the persistence of an uptick in productivity during the COVID-19 period that has subsequently reversed.

Figure 4 - Australian productivity

Source: Reserve Bank of Australia

Past growth has been driven by increases in labour input despite slow productivity growth…

Over the decade prior to COVID-19, New Zealand maintained solid aggregate real GDP growth despite weak labour productivity growth (Figure 5). Could a similar pattern emerge over coming years?

Figure 5 - Real GDP and labour productivity growth

Sources: Stats NZ, Treasury

Over the 2010's New Zealand averaged 3% annual real GDP growth per year. Over this period labour productivity growth averaged 0.4% per annum. Most of the growth in output could be attributed to increased labour inputs (the various grey segments of Figure 6). Strong working age population growth was the major driver contributing on average 1.7 percentage points to growth, four times the contribution of labour productivity.

Figure 6 - Real GDP growth contributors (2009 to 2019)

Sources: Stats NZ, Treasury

Strong growth in labour force participation, including from older workers and women more generally, saw overall rates of participation increase and contributed on average 0.4 percentage points to growth - similar to that of productivity. Aggregate hours worked were also boosted by increases to the number of weekly hours worked by the average worker and declines in the unemployment rate. Together these two factors also made a similar contribution to growth as productivity.

…but there is probably less scope for increases in labour supply to offset weaker productivity growth

We see less scope than in the past for growth in some of the components of labour supply. Population growth will continue to be an important driver of the size of the economy. While we are currently seeing a slowing in net migration, future surges cannot be ruled out. Labour force participation rates remain high but there is a limit to future increases, particularly given the ageing of the population. The upward profile in hours worked per worker that was apparent over the 2010s has reversed in recent years.

Overall, it seems likely that weaker productivity growth will act as a constraint to aggregate GDP growth.

Forecasts will continue to evolve as our understanding of the economy also evolves

Economic data is subject to revision over time, with the timeliest data (eg, quarterly GDP) based on less complete information. Annual data is produced with longer lags but is more comprehensive and a key driver of revisions to earlier quarterly data. Recently Stats NZ released annual nominal GDP data for the year to March 2023. Relative to quarterly data released in September (and used as the base for our forthcoming Half Year Update forecasts) this showed levels of nominal GDP around 1.5% higher.

The release of September 2024 quarter GDP data in December will incorporate this annual data. In addition, provisional data for the March 2024 year will also be included. Revisions to future data therefore have the potential to change our interpretation of recent economic performance, including for productivity. The upward revision for the year to March 2023 includes the current peak in the economic cycle, so revisions beyond that could impact our assessment of the size of the current downturn.

The GDP data released in December will provide an important starting point for our BEFU 2025 forecasts. Ahead of this the Half Year Update will be released on 17 December and will factor in the risks to the outlook discussed in this special topic. In addition to the risks to the economic outlook from lower productivity growth, a recent Treasury speech by Treasury's Chief Economic Advisor on New Zealand's challenging fiscal context highlighted additional risks relating to tax revenue as well as possible areas of focus to lift productivity growth.