Every Budget season in Ireland brings a fresh set of economic forecasts. Growth, inflation, employment, tax revenue — each is projected forward with numbers that look precise to the decimal point. These forecasts are not academic exercises. They determine spending envelopes, guide mortgage holders' expectations and shape business investment decisions. Understanding how they are produced, and where they tend to go wrong, is a basic part of economic literacy.
Who produces Irish economic forecasts
Several institutions produce regular forecasts for the Irish economy. The Department of Finance publishes macroeconomic projections as part of the Budget and the Stability Programme Update. The Economic and Social Research Institute produces quarterly commentary and medium-term scenarios. The Central Bank of Ireland publishes bulletins with forecasts. The European Commission, the OECD and the IMF all include Ireland in their regular outlooks. Each uses different models, different assumptions and different vintages of data.
This plurality is healthy. No single institution has a monopoly on the future, and comparing forecasts across sources is itself informative. When they converge, confidence in the direction is higher. When they diverge, the assumptions driving the differences deserve attention.
The Irish forecasting problem
Ireland is a particularly difficult economy to forecast. The dominance of multinational activity means that headline GDP is volatile and only loosely connected to domestic living standards. A single re-routing of intellectual property or a change in contract manufacturing arrangements can shift GDP by several percentage points. Forecasts that focus on GDP will inevitably be subject to large revisions, and those revisions tell us little about the underlying economy.
The better forecasters have moved towards using modified domestic demand and other adjusted measures as their primary target. This is analytically superior, but it introduces its own challenges — the adjusted measures are themselves estimates, subject to revision, and less timely than the headline figures. There is no clean solution to the fact that the most economically meaningful measures of Irish activity are also the most uncertain.
Where forecasts systematically err
Forecast errors in Ireland cluster around a few predictable areas. Corporation tax revenue is consistently underestimated in expansionary phases, because the concentration of the tax base means that the upside is driven by a handful of firms whose decisions are not visible to forecasters until after the fact. Housing completions are consistently overestimated, because plans and targets do not translate into delivered units at the assumed pace. Inflation has been missed in both directions — too low before the 2022 shock, and too high in the subsequent disinflation.
A forecast is not a prediction of the future. It is a conditional statement about what happens if the assumptions hold. The assumptions rarely hold for long.
The role of judgement
Pure model-based forecasts are rarely published without adjustment. Economists apply judgement — adding or subtracting from model output based on information that the model does not capture. This is necessary, because no model represents the full complexity of the economy. But it also means that forecasts carry the biases and perspectives of the people producing them. An institution that has historically been cautious will tend to produce cautious forecasts; one that has been optimistic will tend in the other direction.
This is not a criticism. It is a fact about how forecasting works. The informed reader of forecasts should know the institutional context of the producer, not just the number at the top of the table.
What to actually trust
The honest answer is that short-term forecasts — one year ahead — carry useful information, while medium-term forecasts should be treated as scenarios rather than predictions. The direction of travel is generally more reliable than the magnitude. A forecast that the economy will grow is more credible than a forecast that it will grow by exactly 3.2 per cent. Ranges and fan charts, where published, are more useful than point estimates because they convey uncertainty honestly.
For practical purposes — Budget planning, mortgage decisions, business investment — the best approach is to look at multiple forecasters, understand the assumptions behind them, and plan for a range of outcomes rather than a single number. The forecast that turns out to be correct is rarely the one that was most confidently presented. The lesson is not to abandon forecasting, but to read it with the scepticism it deserves.
