Chapter 8 of John Doyle’s Before the Vote: Planning for a united Ireland argues that economic convergence between North and South is “the most likely result of Irish unity”, and that Northern Ireland would be paying its way within a decade. The chapter contains three kinds of claim. Some changes would follow from unity itself. Others are policy choices that Doyle says Stormont has failed to make and that do not, in principle, require unity. The growth projections that carry the fiscal conclusion are assumptions borrowed from Central European economies that started much poorer, and the tables reach balance partly because they hold most public spending flat for ten years.
What unity would change
Doyle lists the changes that would come on the first day: EU membership, full access to the single market for goods and services, the euro, Ireland’s visa rules and the Republic’s corporation tax regime, and the Republic’s reputation as a location for foreign investment. Some of this is already partly in place. Doyle accepts that the Windsor Framework gives Northern Ireland “some advantages for exports in goods”, so the main new gain in market access would be in services, where he says Northern Ireland is excluded and where its exports are weakest. The corporation tax rate would be a policy choice, and its revenue cost does not appear in his tables.
Many of the other reforms in the chapter do not depend on unity. Doyle wants the transfer test abolished, a programme like DEIS for disadvantaged schools, an expanded university in the North West and joint planning of roads, rail and energy. He describes these as decisions “the local political system there has been incapable of taking”, and says of cross-border infrastructure that it “could and should be done now”. His case for linking them to unity is that Stormont has not delivered them. That is an argument about governance, and the chapter does not show that a united Ireland’s institutions would make these decisions more readily.
Trade with Britain
Doyle relies on Seamus McGuinness and Adele Bergin, who found that sales to Britain were 20% of Northern Ireland’s business turnover in 2016, and concludes that disruption to trade with Britain “is likely to be at a very low level”. The share of turnover is similar now. NISRA’s trade statistics for 2024 put sales to Britain at £20.1 billion, 18.4% of total sales. But that figure is larger than all exports to the Republic, the rest of the EU and the rest of the world combined, which come to £19.6 billion. Northern Ireland also bought £16.4 billion of goods and services from Britain, four times its £4.1 billion of purchases from the Republic.
Northern Ireland already applies EU customs rules to goods arriving from Britain, but the Windsor Framework provides simplified procedures for goods that stay in Northern Ireland, and goods from Northern Ireland enter Britain without checks. Unity would remove Northern Ireland from the UK internal market, and without new arrangements, trade with Britain would operate under the rules that apply to trade between Britain and the Republic. Doyle expects trade to continue on a “no tariffs, no quotas” basis. FitzGerald and Morgenroth, in their 2024 note for the IIEA, expect separation from the UK economy to “involve major costs”. Neither puts a figure on it, and a share of turnover does not measure the cost of new checks on the largest external market and the main supplier.
The Central European comparison
Doyle’s three growth paths are based on the experience of states that joined the EU from 2004 onwards. He says Latvia, Lithuania, Bulgaria and Romania grew about 3 percentage points a year faster than the EU average, a middle group about 2 points faster, and Czechia, Hungary and Slovenia about 1 point faster. Eurostat’s figures for real GDP per head, measured from each country’s year of accession to 2023, give a looser pattern. Lithuania and Poland grew about 2.7 to 2.9 points a year faster than the EU average, Romania, Croatia and Bulgaria about 2.2 to 2.4, Latvia and Slovakia about 2, Estonia and Hungary about 1.1 to 1.2, and Czechia and Slovenia less than 1. Poland belongs in Doyle’s top group, Estonia in his bottom one, and Latvia, Romania and Bulgaria in the middle.
The same figures suggest a more important qualification. The fastest-growing economies generally started from much lower output per head, although the comparison uses GDP in euros, not purchasing-power levels. In 2004, output per head in Lithuania, Latvia, Poland, Romania and Bulgaria was between a fifth and a third of the EU average, while Czechia and Slovenia, at over half, grew most slowly. Doyle does not compare Northern Ireland’s starting position with those of the others. He calls paths A and B “the most likely outcomes in practice”, but the chapter offers no model of Northern Ireland’s economy to support that ranking.
How the tables reach balance
Tables 8.1 to 8.3 can be reproduced. They repeat Tables 4 to 6 of a report Doyle wrote in 2025 for a joint research programme of Dublin City University and the Ulster University Economic Policy Centre, and he published them again in Irish Studies in International Affairs in 2026, with the same figures in every cell. Tax revenue grows each year at Northern Ireland’s past rate of 1.3%, plus the assumed boost of 1, 2 or 3 points, plus a small amount from population growth. Spending rises only by the €1 billion added in year one, the cost of raising public pay over 15 years and the gradual transfer of pensions. The notes say that all other policy and tax decisions “are held constant”, so apart from the specified additions, aggregate public spending stays flat in real terms for ten years. The tables do not model the costs that an ageing population, rising demand for health care or other pressures would add. On that method, even with no boost from unity, the deficit would fall from €3.0 billion in year one to about €2.3 billion in year 10. Part of the improvement in every scenario therefore comes from growth Northern Ireland would have anyway and from a decade of flat aggregate spending.
The tables also inherit the starting deficit from Chapter 1. That figure assumes the UK would keep paying existing pensions and that Ireland would take on no share of UK debt costs, and its pension deduction needs correcting in two respects. If the Treasury’s old-age category is the basis of its state pension figure, as the numbers indicate, it subtracts £628 million of care services for older people as if they were pensions. That figure is also for 2021/22, a year earlier than the deficit; once the years align, the deduction is £360 million too high on the Treasury’s own pensions line, and £1,284 million too high if it covers only the State Pension, the contributory benefit on which Doyle’s argument rests. The £1.7 billion for public service pensions is Northern Ireland’s population share of a Treasury measure that includes the annual growth in pension liabilities, which the regional deficit does not count. The Treasury’s regional data record £224 million of public service pension spending in Northern Ireland in 2022/23. Doyle describes the cost as the Treasury’s “top-up” between contributions collected and pensions paid, and on that description the deduction is £1,484 million too high. Under the settlement he proposes, in which the UK pays pensions already earned, the right figure would depend instead on the gross pensions paid, which the published data do not show.
As a sensitivity test, the smaller state pension correction, about €424 million at Doyle’s exchange rate, delays balance on the low and middle paths by a year, to year 10 and year 7, and leaves the high path in surplus in year 5. The larger one, about €1.5 billion, delays balance to year 6 on the high path, year 8 on the middle path and year 12 on the low path, which is still €729 million in deficit in year 10. The public service pension correction alone, about €1.75 billion, gives the same balance years, with the low path €964 million in deficit in year 10. Combined with the smaller state pension correction, the balance comes in year 7, year 9 and year 13 on the high, middle and low paths. Combined with the larger one, it comes in year 8, year 10 and year 15, and the low path is still €2.5 billion in deficit in year 10. Doyle’s worst case, if the UK paid no pensions, adds “approximately £4 billion” to the deficit, while Chapter 1 gives the pension cost as £5.7 billion for 2022/23 and £4.4 billion after an expected fall that appears to rest on the same Treasury measure of public service pension costs.
A second model
Fine Gael published A Blueprint for a Unified Island at its Ard Fheis in Cork on 7 November, with an economic appendix by Stephen Kinsella of the University of Limerick. Kinsella starts from Doyle’s adjustments to the published deficit and arrives at the same inherited deficit, €1.78 billion in 2023 euros, if the UK keeps existing pensions and debt interest. His model therefore inherits the pension figures questioned above. If the UK keeps existing pensions, the corrections set out there would raise the deficit his model starts from; their effect under the other settlements has not been calculated. Unlike Doyle’s tables, Kinsella’s model lets public pay, benefits and purchases grow by 1% a year in real terms and with population, and it tests the result at a range of growth rates instead of Doyle’s three paths, under which tax revenue grows by about 2.5% to 4.5% a year.
Kinsella reports four estimates of the cost of raising pay and welfare to the Republic’s levels and treats none as central. The closest estimate to Doyle’s assumptions uses Doyle’s pay gap and his costing of pension equalisation, and Kinsella notes it draws on two studies that find unity more affordable than some critics claim. Under the design that phases pay and welfare over 15 years, the period Doyle uses for public pay, that case costs €4.0 billion in year one. At 2% growth in Northern Ireland, the yearly cost is still €5.3 billion in year 25, and repaying the transition borrowing would need growth of about 3.2% a year. At Northern Ireland’s average of 1.3% since 2000, the cost in year 25 is €12.1 billion. Removing the extra investment the design assumes still leaves €4.8 billion in year 25 at 2% growth. In the two cases that use FitzGerald and Morgenroth’s larger estimate of the welfare gap, the cost in year 25 at 2% growth is €17.8 billion and €24.1 billion.
The two models make different assumptions, and Kinsella’s results leave the arithmetic of Doyle’s tables intact. They show that, in a model where spending grows, the case closest to Doyle’s own assumptions still imposes an additional yearly cost on the Republic’s Exchequer after 25 years, at a growth rate above Northern Ireland’s average since 2000.
The case for convergence
Doyle’s strongest point is the evidence that Northern Ireland has fallen further behind the Republic since 1998. The ESRI study he relies on finds productivity per worker in the Republic about 40% higher by 2020, even after excluding the sectors dominated by multinationals. Elsewhere, he restates this as Northern Ireland being “over 40 per cent lower”, which overstates the gap: a level 40% higher in the Republic leaves Northern Ireland about 29% lower. He argues that there is “no evidence” the trend will change while Northern Ireland remains in the UK, although the evidence he presents does not establish that it could not. His projections show what would follow if Northern Ireland grew like the faster accession states while most spending stayed flat, but not that it would. Before they can carry the weight the chapter places on them, they must separate the gains from EU membership itself from those that depend on new policies, and explain why Northern Ireland would grow well above the 1.3% a year it has averaged since 2000.
Sources:
John Doyle, Before the Vote: Planning for a united Ireland, Dublin: DCU Press, 2026 (open access): pp. 38 (ESRI productivity comparison, multinational-dominated sectors excluded); 161 (“over 40 per cent lower”; “no evidence”); 173 (“40 per cent higher”); 162–3 (changes on the first day); 164–5 (education; “incapable of taking”); 167 (Windsor Framework, “some advantages for exports in goods”; services); 169–70 (McGuinness and Bergin; “no tariffs, no quotas”; “a very low level”); 171 (“could and should be done now”); 175–6 (growth paths A to C; 1.3% growth, 2000–2024); 177 (“the most likely outcomes in practice”); 178–81 (Tables 8.1 to 8.3 and notes); 182 (worst case, “approximately £4 billion”; “the most likely result of Irish unity”); and for Chapter 1, pp. 16 (£1.7 billion for public service pensions; Treasury “top-up”), 28 and 32 (£5.7 billion and £4.4 billion) and 34, note 1 (“a per capita breakdown of the UK figure”).
NISRA, Northern Ireland Economic Trade Statistics 2024, 11 March 2026 (provisional; total sales £109.3 billion; sales to GB £20.1 billion; exports to Ireland £10.5 billion, rest of EU £3.6 billion, rest of world £5.5 billion; purchases from GB £16.4 billion, from Ireland £4.1 billion).
John FitzGerald and Edgar Morgenroth, Northern Ireland Subvention: Possible Unification Effects, IIEA, April 2024, section 1 (“major costs”).
John Doyle, The Projected Public Finances of the Early Years of a United Ireland, and the Northern Ireland Subvention, Dublin City University and Ulster University Economic Policy Centre (All-Island Economy Project), 2025, Tables 4–6, pp. 28–9. John Doyle, “Projected Public Finances of the Early Years of a United Ireland, and the Northern Ireland Subvention”, Irish Studies in International Affairs, 37(2), 2026, pp. 130–64, Tables 5–7, pp. 160–1.
Fine Gael Northern Ireland Engagement Group, A Blueprint for a Unified Island, Fine Gael, published at the 83rd Ard Fheis, Cork, 7 November 2026. Stephen Kinsella, The Economics of Reunification, Appendix B to the Blueprint: pp. 2 (four harmonisation cases, none central; the rate-based case draws on “two studies”), 11 (inherited deficit of €1.78 billion), 23 (Table 8, design settings), 24 (Doyle’s adjustments; spending growth of 1% a year in real terms, scaled by population), 28 (Table 9; average growth of 1.3%, 2000–2024) and 30–1 (Table 12, growth of 1.3% and no extra investment).
Eurostat, Real GDP per capita (sdg_08_10), chain-linked volumes (2020), euro per capita (data last updated 22 October 2026, extracted 25 October 2026). Growth above the EU average is calculated as the difference in compound annual growth rates from each country’s year of accession to 2023: 2004 for most, 2007 for Bulgaria and Romania, and 2013 for Croatia. Starting levels relative to the EU average are at 2020 prices in euros, not purchasing power standards.
HM Treasury, Public Expenditure Statistical Analyses 2023, paras D.18–D.19 and Table D.1, p. 220 (net public service pensions, GAAP basis, £58,908 million in 2021/22; changes in liability removed to reach the National Accounts basis). HM Treasury, Country and Regional Analysis 2023 database, December 2023, Northern Ireland, sub-functions 10.2 and 10.3, 2022/23 (Northern Ireland’s own public service pension schemes, segments S099A063–S099A067, £149.3 million; Northern Ireland’s share of UK-wide public service pension schemes, £74.3 million). NISRA, 2021 Mid-Year Population Estimates for Northern Ireland (1,904,600). Office for National Statistics, Population estimates for the UK, England, Wales, Scotland and Northern Ireland: mid-2021 (67,026,292).
Calculations from Doyle’s tables: tax revenue grows at 4.46%, 3.46% and 2.46% a year on paths A, B and C (from €25,367 million in year one); at 1.3% alone the year-ten balance is −€2,293 million; the pension correction of £360 million is £4,014 million (HM Treasury, Public Expenditure Statistical Analyses 2023, Table 10.4) less £3,654 million for pensions alone in 2022/23 (HM Treasury, Public Expenditure Statistical Analyses 2024, Table 10.4, p. 172), converted at Doyle’s rate of €1,750 million to £1,487 million (about €424 million); the State Pension-only correction of £1,284 million is £4,014 million less £2,730 million (Department for Communities, Annual Report and Accounts 2022–23, HC 1601, Note 4b, p. 163), about €1,511 million; the public service pension correction of £1,484 million is £1,708 million (£5,722 million less £4,014 million) less £224 million (£223.6 million in the Treasury’s regional data for 2022/23), about €1,746 million, and the population share of the UK measure is £58,908 million × 1,904,600/67,026,292 = £1,674 million; the combined corrections are about €2,170 million and €3,258 million; balance years recalculated from Doyle’s tables with revenue growing as above and expenditure rising by €267 million a year after year one, with Doyle’s yearly additions for pay and pensions unchanged; 1 ÷ 1.4 = 0.71, so a level 40% higher in the Republic leaves Northern Ireland about 29% lower.
El Cavador is a Slugger reader from Belfast with a particular interest in education
Discover more from Slugger O'Toole
Subscribe to get the latest posts sent to your email.
