Producing Short-Term Forecasts of the Irish Economy: A Suite of Models Approach.

A new working paper from Niall Conroy and Eddie Casey of the Fiscal Council Secretariat.

Abstract:

The Council’s mandate includes endorsing, as it considers appropriate, the official macroeconomic forecasts of the Department of Finance on which the annual Budget and Stability Programme Update are based. As part of the endorsement process and for the purposes of its ongoing monitoring and analysis of the Irish economy, the Council’s Secretariat produces its own Benchmark macroeconomic projections. This paper describes the short-run forecasting models used by the Secretariat for producing these projections. The general forecasting approach can be described as follows. Equations are used to forecast each component of the expenditure side of the Quarterly National Accounts. Multiple models are estimated for most components, with the simple model average used as an initial input into the formulation of the Benchmark projections. The out-of-sample forecasting performance of these models is assessed at each endorsement round. In addition to these model-based projections, other elements are considered. Discussions with the Council and other forecasting agencies help to guide any judgement that may be applied before arriving at the final Benchmark projections.

4 replies on “Producing Short-Term Forecasts of the Irish Economy: A Suite of Models Approach.”

Why doesn’t the Department of Finance publish forecasts for who will win the Grand National next year? They would probably be more accurate, a better use of their taxpayer-funded resources. Future economic growth is unforecastable. Why don’t they simply accept that and stop trying to play God? Government should focus on making the economy as nimble and fit as possible, with taxes and government spending as low as possible, so as to cope with whatever opportunities and/or shocks the future throws up, without trying to forecast what they will be. The past record of the main economic forecasters in Ireland is abysmal. This is particularly true of their record in forecasting population growth. The current housing shortage is due to population growth running far ahead of what the forecasters said it would be just a few years ago.

Economic and demographic forecasts are often influenced by the political climate and lack any statistical basis. A case in point is Brexit. Because the political consensus in Ireland is virulently hostile to Brexit, nearly all forecasts predict only doom and gloom resulting from Brexit. I myself voted in the referendum to remain, but I’ve never swallowed the line that Brexit will lead to economic catastrophe. Since the referendum last June unemployment in Ireland has fallen at well over twice the rate it fell in the preceding 12 months. Who forecast that? If I recall correctly, the consensus forecast was that there would be an immediate downturn following the referendum if Brexit won because of the uncertainty that would induce. It simply hasn’t happened either in Ireland or Britain.

As the authors note, quarterly Irish GDP data is very volatile and more prone to substantial revisions than other developed economies. In other economies, such as the EA, the relationship between indicators such as PMI’s and quarterly GDP is fairly close, but not so for Ireland.
On the detail it is interesting that credit growth or the rate of interest does not feature in any of the equations, (although the stock of household debt does impact net wealth and hence consumption) and financial variables and conditions in general have no role to play. Also the (real) exchange rate has no impact on exports volumes ( there is not much empirical support for that in international studies either).

Comments are closed.