Esmond Birnie: There’s no evidence that having Stormont creates an economic dividend for Northern Ireland
The hard evidence for this proposition is less obvious.
Take one of the key summary statistics of economics performance: the size of the Northern Ireland economy divided by the population compared to the UK average, ie Gross Value Added per person [Note 1, below].
The Office for National Statistics last week provided its update of GVA per person levels for the UK region. A graph shows how NI compared to the UK average during 1998-2020 (similar figures for Scotland are also included).
Advertisement
Hide AdAdvertisement
Hide AdThere is some year to year volatility in NI’s comparative position but a number of notable conclusions arise:
There is no clear evidence of a strong catching up process whereby during the 1998-2020 period NI would have narrowed the gap between NI and UK average levels of output per person.
Remember, 1998 was the year of the Belfast Agreement. Devolution occurred at the end of 1999 (continued fitfully through to 2002, then suspended, restored in 2007 but collapsing in January 2017, restored three years later).
In 1999 NI’s comparative level was 79.5% and in 2020 it was 79.7%. In other words, practically the same.
Advertisement
Hide AdAdvertisement
Hide AdDuring the (relatively) ‘long’ period of devolution 2007-2017 NI’s comparative position actually got worse (80.8% down to 77.1%). Conversely, there was a slight improvement during the following three years of caretaker management by the senior civil servants.
[Source: Dataset to ONS 30 May 2022, “Regional economic activity by GDP, UK: 1998-2020”]
None of this is say that devolution isn’t desirable: it may well be desirable in political terms especially in terms of increasing accountability.
It is just to note that the evidence to date doesn’t include any strong suggestion of a large economic dividend.
Advertisement
Hide AdAdvertisement
Hide AdIt is perhaps not much consolation to Northern Ireland to note that there may be parallels in the Scottish experience where there has been continuous devolution since 1999.
It is true that Scotland’s comparative position in 2020 (92% of the UK average) is up relative to 1999 (89.1%). However, Scotland’s comparative position actually worsened during the final 11 years of the data: from 96.1% in 2009 to 92% in 2020.
One further observation based on recent ONS data, on 7 June ONS published its most recent ‘Model-based early estimate of regional GVA’ for the UK regions and this related to growth in the first 3 months of 2022.
This ONS data attempts to provide really up-to-date measures of economic growth but is subject to many caveats [Note 2].
Advertisement
Hide AdAdvertisement
Hide AdFor what it is worth, the ONS data shows that in the First Quarter 2022 (ie January-March) the NI economy grew by 0.4% or half the UK average of 0.8%.
In the past some of this ONS ‘nowcasting’ type data has been used to support the claim “The NI economy is out-growing the rest of the average and this is because of the impact of the Protocol”.
Now that the ONS data is showing a much slower growth rate in NI will we hear the reverse argument?”
• Dr Esmond Birnie, Senior Economist Ulster University Business School
• Notes:
Advertisement
Hide AdAdvertisement
Hide Ad1. GVA differs from the more “popular” measure GDP but is a close approximation. To go from GVA to GDP add taxes on products and subtract subsidies on products. A deeper criticism of too much reliance on GVA or GDP statistics is that such data does not sufficiently allow for all the factors that impact on well-being or quality of life whether negatively or positively. That said, GDP/GVA statistics have been with us since the 1950s whether measures of happiness/well-being are a much more recent development and there are questions about how far (personal self-assessments) of well-being/happiness are truly comparable across regions or countries or across time. In any case, there is much evidence that there is substantial correlation between the level of GDP per person and a range of indicators of social progress, e.g. educational and health standards.
2. It remains experimental data. To some extent it is based on projecting forward relationships between regional growth figures based on historical data. In the NI case there is strong reliance on the NI Composite Economic Index which (notably) relies on a job count to proxy for public sector output change. The Margins of error are very considerable