On 11 February 2026, Transparency International published the Corruption Perceptions Index 2025. Ireland scored 76 out of 100 and ranked twelfth out of one hundred and eighty-two countries. The Department of Foreign Affairs, the IDA Ireland investor materials, the embassies, all carried the result through the spring. The same year, on a different timeline, Ireland sat ninth of seventy jurisdictions on the Tax Justice Network's Corporate Tax Haven Index, with a Haven Score of 78 and an estimated 2.9% of all global multinational financial activity routing through the country. Two reputable international indices, applied to the same country in the same year, returned diametrically opposite assessments. The CPI placed Ireland near the top of the cleanest table. The CTHI placed Ireland near the top of the most permissive. Both numbers are technically accurate. The disagreement is the diagnosis.

The question this piece sets out to answer is which index is asking the right question, and what the answer says about how the international architecture of measurement participates in the form of governance the recent pieces on this site have named as acceptable corruption.

The two indices side by side

Both indices come from reputable international research bodies. Both use published methodology. Neither is fringe.

The Corruption Perceptions Index is published annually by Transparency International, the Berlin-based anti-corruption NGO founded in 1993. The 2025 edition aggregates thirteen data sources from twelve institutions, covering data published in the previous two years. The included sources are the World Bank Country Policy and Institutional Assessment, the World Economic Forum Executive Opinion Survey, the Bertelsmann Foundation Transformation Index, the Economist Intelligence Unit Country Risk Service, Freedom House Nations in Transit, the Global Insight Country Risk Ratings, the IMD World Competitiveness Yearbook, the PRS International Country Risk Guide, Varieties of Democracy, the World Justice Project Rule of Law Index, the PERC Asia Risk Guide, the Bertelsmann Sustainable Governance Indicators, and the African Development Bank Country Policy and Institutional Assessment.

What the underlying surveys ask experts and business respondents to assess: bribery in the awarding of contracts, diversion of public funds, use of public office for private gain, nepotism in civil-service appointments, state capture and the executive's ability and willingness to enforce integrity mechanisms.

What Transparency International explicitly acknowledges the CPI does not measure: private-sector corruption, tax fraud, money laundering and citizens' direct experiences. The CPI 2024 publication page carries the caveat directly: "While the CPI does not measure this, dirty money poses a major corruption problem with harmful effects that reach far beyond these countries' borders." TI is internally honest about the scope. The honesty does not survive the headline use.

The Corporate Tax Haven Index is published biennially by the Tax Justice Network, the London-based research organisation founded in 2003 to address international tax avoidance and financial secrecy. The CTHI measures the extent to which each jurisdiction enables corporate tax avoidance globally, combining a Haven Score (how aggressively the jurisdiction's laws and regulations facilitate avoidance) with a Global Scale Weight (how much of global multinational financial activity uses the jurisdiction).

Ireland's 2024 CTHI position: Haven Score 78, CTHI value 1,432, ninth of seventy jurisdictions. The Tax Justice Network estimates that 2.9% of all global multinational financial activity routes through Ireland. The top ten on CTHI 2024 in order: British Virgin Islands, Cayman Islands, Switzerland, Bermuda, Singapore, Hong Kong, the Netherlands, Jersey, Ireland, Luxembourg.

The Financial Secrecy Index, also published by the Tax Justice Network, measures jurisdictions' contribution to global financial secrecy. Ireland's most recent FSI position is approximately fourteenth globally, in the upper-middle of the table.

Two instruments, two verdicts. One places Ireland near the top of the cleanest. The other places Ireland near the top of the most permissive. The contradiction is not noise. It is the structural finding the indices return when read together.

The GDP and GNI star parallel

The Irish state has been through this question before, on the economic-measurement layer, and the answer is on the public record.

On 12 July 2016, the Central Statistics Office published its revised national accounts for 2015. The headline figure was Irish GDP growth of 26.3%, revised upward from an initial estimate of approximately 7.8%. Twenty-four minutes after the CSO release, at 11:24 GMT, Paul Krugman posted on Twitter: "Leprechaun economics: Ireland reports 26 percent growth! But it doesn't make sense. Why are these in GDP?"

The figure was real. It was also meaningless as a description of the Irish economy. Subsequent research by Seamus Coffey, then chair of the Irish Fiscal Advisory Council, and by Brad Setser at the Council on Foreign Relations identified Apple's first-quarter 2015 onshoring of intellectual property via the Capital Allowances for Intangible Assets relief, the structure later known as the "Green Jersey," as the dominant single driver of the jump. Other multinational balance-sheet relocations contributed. The Irish statistical apparatus had recorded as Irish economic output a series of intra-corporate transfers that did not correspond to any meaningful production or employment in the country.

The political response was the Economic Statistics Review Group, established by the CSO and chaired by then-Central Bank Governor Philip Lane. The group's central recommendation was a new measure: modified Gross National Income, written GNI, designed to strip out the categories of multinational accounting that were producing the distortion. GNI equals GNI minus three categories: depreciation on foreign-owned intellectual-property assets, depreciation on aircraft-leasing company assets, and the net factor income of redomiciled public limited companies.

GNI was first formally published by the CSO in July 2017, in the National Income and Expenditure 2016 release. It has been published alongside GDP in every subsequent annual release. The 2024 figures, published in 2025, illustrate the persistent gap. Irish GDP for 2024 was €562.8 billion. Irish GNI for the same year was €321.1 billion. The gap is €241.7 billion. GNI* is 57.1% of GDP. The Irish statistical authority itself, by its own publication choices, does not consider headline GDP a meaningful measure of Irish economic activity. It publishes a parallel measure specifically because the headline number is known to be distorted.

The international architecture quotes the headline number anyway. World Bank data tables, IMF databases, OECD comparisons, sovereign credit ratings, EU budget contributions, GDP-per-capita rankings that place Ireland among the wealthiest countries in the world, all use the GDP figure that the Irish state itself knows is wrong. The Irish state pays its EU contributions and accepts its rankings on the basis of a number its statistical authority has publicly disowned as a measure of the country.

The CPI is structurally the same instrument at the corruption layer. It measures what corruption looks like from outside the country, viewed through the lens of bribery and prosecutable embezzlement. It does not measure what corruption looks like from inside the country, performed through statute, through administrative ruling, through regulatory capture, through the engineered holes the recent piece on this site set out. The international architecture quotes the CPI anyway. The Department of Foreign Affairs cites the rank. The IDA cites the rank. The embassies cite the rank. Foreign-investment pitches cite the rank. The rank is technically accurate against what it measures. The thing it measures is not the form of corruption the country practises.

The parallel is exact. GDP is to economic activity what CPI is to corruption. Both metrics measure something. Neither metric measures the country.

The comparative cases

The pattern is not Irish exceptionalism.

Switzerland scored 80 out of 100 and ranked seventh on the CPI 2025. On the CTHI 2024 Switzerland ranks third, behind only the British Virgin Islands and the Cayman Islands. Switzerland is, by the same year and by reputable measurement, in the global top ten on both the cleanest-public-sector and the dirtiest-tax-haven indices.

Luxembourg scored 78 and ranked eighth on the CPI 2025. On the CTHI 2024 Luxembourg ranks tenth. Same pattern.

Singapore scored 84 and ranked third on the CPI 2025, holding its position from the previous year. On the CTHI 2024 Singapore ranks fifth. Same pattern.

Ireland scored 76 and ranked twelfth on the CPI 2025. On the CTHI 2024 Ireland ranks ninth. Same pattern.

Four jurisdictions, in the same publication cycle, in the global top fifteen on both indices simultaneously. The form of corruption these jurisdictions practise is structural, performed at the corporate-international layer through statute and administrative ruling, and systematically invisible to the CPI methodology because that methodology was designed to detect bribery and prosecutable embezzlement at the public-sector layer. The CPI does not have a Switzerland problem or an Ireland problem. The CPI has a structural-corruption blindness when used as a headline measure of governance integrity.

The political economy of measurement

The argument that metrics carry political function, rather than serving as neutral technical instruments, has a substantial scholarly backbone.

Theodore Porter's Trust in Numbers: The Pursuit of Objectivity in Science and Public Life (Princeton University Press, 1995) is the foundational text. Porter argues that quantification is a "technology of distance," adopted not because numbers are inherently more accurate but because they substitute mechanical objectivity for personal judgement in low-trust institutional settings. Quantification is a defensive political strategy of professionals exposed to outside scrutiny, not a neutral epistemic advance. The CPI, like GDP, is exactly this kind of instrument: a quantitative answer that allows a political question to be deferred from substantive contest to methodological adjudication.

Diane Coyle's GDP: A Brief but Affectionate History (Princeton University Press, 2014) traces GDP from its 1930s war-economy origins under Simon Kuznets and Richard Stone. Coyle's central argument is that GDP was built for an industrial-manufacturing economy of the mid-twentieth century and struggles with the services economy, the digital economy, unpaid household production, environmental depletion and quality change. The Irish GDP / GNI* divergence is one of the cleanest live demonstrations of Coyle's case. The measurement instrument designed for one economy is being used to describe another.

Lorenzo Fioramonti's Gross Domestic Problem: The Politics Behind the World's Most Powerful Number (Zed Books, 2013) sharpens the political-economy reading. Fioramonti argues that GDP is not a neutral statistic but a political instrument that emerged with and sustains a particular growth-centred power structure. Entrenched interests, including finance, ratings agencies, and growth-coalition states, defend GDP against alternatives despite known methodological defects. The defence is structural, not technical.

Mariana Mazzucato's The Value of Everything: Making and Taking in the Global Economy (Allen Lane, 2018) addresses the same problem at the level of the production boundary. Mazzucato argues that the boundary inside national accounts has been quietly redrawn over recent decades so that finance, rent extraction and monopoly pricing now register as value creation rather than value capture. GDP therefore rewards extractors and obscures the public sector's productive role. The argument applies directly to Ireland. The corporate financial flows that produced the 26.3% GDP jump in 2015 are recorded as value creation in the national accounts.

Mary Poovey's A History of the Modern Fact: Problems of Knowledge in the Sciences of Wealth and Society (University of Chicago Press, 1998) makes the deeper historical case. Poovey argues that the "modern fact" as a separable, theory-neutral unit of knowledge is itself a historical construction, traceable from double-entry bookkeeping in the late sixteenth century to the institutionalisation of statistics in the 1830s. Numerical representation acquired epistemic authority because it served governmental and commercial needs, not because numbers are neutral. The CPI and GDP are inheritors of that history. Their authority rests on what they let governmental and commercial actors do, not on what they describe.

On the CPI specifically, the foundational political-science critique is Staffan Andersson and Paul M. Heywood, "The Politics of Perception: Use and Abuse of Transparency International's Approach to Measuring Corruption," in Political Studies volume 57, number 4 (2009), pages 746 to 767. Andersson and Heywood accept the CPI's analytical contribution while attacking its headline use. Their critique covers conceptual narrowness in the CPI's definition of corruption, methodological weakness in aggregating heterogeneous surveys, false precision in country rankings, and a policy danger they name the "corruption trap," in which aid conditionality demands reforms that are impossible without aid. Heywood has continued the line of work through the Sussex Centre for the Study of Corruption.

The counter-argument

The defenders' case is not weak and should be named honestly. Transparency International publishes the CPI's limits in its own documentation. The 2024 page carries the dirty-money caveat directly. The methodology has had substantial improvements over thirty years, including the 2012 revision that moved the index to a 0-to-100 scale, established the four-stage aggregation process and made year-on-year comparison possible from 2012 onwards. The CPI as Transparency International publishes it is internally honest. The defenders' point that no single index can do everything is true. The critique offered here is not that the CPI is dishonest at the methodological layer. The critique is that the CPI is misused at the rhetorical and political layer, and that the misuse is structural rather than accidental.

The Sustainable Development Goal 16 indicators offer one example of what a different methodology can do. Indicator 16.5.1 measures the share of persons who, in contact with a public official in the previous twelve months, paid or were asked for a bribe. Indicator 16.5.2 is the same for firms. The custodian agency is the UN Office on Drugs and Crime. The data is experiential rather than perception-based, collected through standardised population and enterprise surveys. The methodology has different problems, primarily coverage gaps and the bribery-only scope, but it does not have the perception-of-cleanliness problem the CPI has. SDG 16.5 exists. It does not get cited.

The OECD Better Life Index, launched in May 2011 as a direct response to the 2009 Stiglitz / Sen / Fitoussi Commission on the Measurement of Economic Performance and Social Progress, is the equivalent move at the economic-measurement layer. Eleven dimensions, user-weightable, an explicit attempt to capture quality-of-life and distributional aspects that GDP averages over. The New Zealand Treasury's Living Standards Framework and the first Wellbeing Budget delivered on 30 May 2019 went further: a national budget formally structured around twelve wellbeing domains and four capitals. Both exist. Both are methodologically robust. Neither has displaced GDP in international comparison.

The alternatives have been built. The reason they are not adopted is political-economy, not technical.

The empirical asymmetry in Irish coverage

The Corruption Perceptions Index release attracts dedicated same-day or next-day news coverage in Irish mainstream media. RTÉ ran "Ireland unchanged at 10th in global corruption index" on 11 February 2025 covering the CPI 2024. The Irish Times ran "Ireland enters top 10 of countries seen as having least public corruption" in January 2023 when Ireland first cracked the top ten. The pattern is uniform across the CPI 2024 and CPI 2025 release cycles. Transparency Ireland's own press release is picked up. The result is framed positively around Ireland's rank.

The Corporate Tax Haven Index 2024 release attracted no dedicated coverage in site-restricted searches of RTÉ, Irish Times and The Journal. The Financial Secrecy Index releases of recent years follow the same pattern. The most prominent Irish Times piece on the FSI, "Ireland steadily climbs financial secrecy rankings," dates to February 2018, when Ireland reached twenty-sixth. Subsequent FSI releases have not surfaced as discrete Irish news stories at the same outlets. Where TJN data is referenced in Irish coverage it is typically embedded in commentary pieces, academic explainers, or coverage of external critics, not framed around the index release as a news event.

The contrast is not bias in the individual journalists. It is the structural alignment of press incentives with the apparatus the previous pieces on this site have set out. The instrument that flatters the country gets dedicated standalone news coverage. The instruments that condemn the country at the same time, on the same kind of evidence base, do not.

The structural close

The headline metric is the wrong metric for what it is asked to measure. It is the right metric for what it is used to do, which is to legitimise a country that practises a form of corruption invisible to its methodology. The same is true of GDP. The same is true of the way "Programme for Government" commitments are quoted as policy, the way "carbon budgets" are quoted as climate commitment, the way "lobbying register" is quoted as transparency, the way "tribunal findings" are quoted as accountability. The metric is the dressing. The substantive measure is held off in alternative instruments that exist, are methodologically robust, and are not adopted because adopting them would name what the country is.

This is the wealth-not-work axis operating at the measurement layer. Headline measures of national success are constructed in the direction of capital. Headline measures of national governance are constructed in the direction of the capital-aligned political class. The Stiglitz Commission report, the OECD Better Life Index, the New Zealand Living Standards Framework, the Tax Justice Network's CTHI and FSI, the SDG 16.5 bribery indicators, the Andersson and Heywood critique, all of these exist. None of them displace the headline instruments. The reason is not technical capability. The reason is who benefits from which measurement.

Two indices, one country, opposite verdicts. CPI 76 out of 100, rank 12. CTHI Haven Score 78, rank 9. Both numbers are technically accurate against what they ask. The CPI is the right number for the wrong question. The CTHI is the right number for the right question. The country quotes the first and ignores the second. The international architecture quotes the first and overlooks the second. The four pieces published on this site over the past week describe the operating system the country runs on. The CPI ranking is one of the legitimising instruments that lets the operating system run.

The standard is not being met because the standard is not measuring the thing. The metric is not the wrong metric by accident. It is the wrong metric by design, and the design is the diagnosis.

Source notes. Transparency International, Corruption Perceptions Index 2024 and 2025, transparency.org. CPI 2024 score 77/100 rank 10/180 confirmed by RTÉ news 11 February 2025; CPI 2025 score 76/100 rank 12/182 published February 2026 via transparency.ie and transparency.org. Tax Justice Network, Corporate Tax Haven Index 2024, cthi.taxjustice.net; Ireland Haven Score 78, rank 9 of 70, 2.9% of global multinational financial activity. Tax Justice Network, Financial Secrecy Index, fsi.taxjustice.net. CSO, National Income and Expenditure 2016 (July 2017 release, first formal GNI publication); CSO Annual National Accounts 2024 (GDP €562.8bn, GNI €321.1bn). Paul Krugman, "Leprechaun economics" post, Twitter, 12 July 2016, 11:24 GMT. Seamus Coffey and Brad Setser identified Apple's Q1 2015 Capital Allowances for Intangible Assets onshoring as the dominant driver. Economic Statistics Review Group, chaired by Philip Lane, 2016-2017. Commission on the Measurement of Economic Performance and Social Progress (Stiglitz, Sen, Fitoussi), final report September 2009. Theodore M. Porter, Trust in Numbers, Princeton University Press, 1995. Diane Coyle, GDP: A Brief but Affectionate History, Princeton University Press, 2014. Lorenzo Fioramonti, Gross Domestic Problem, Zed Books, 2013. Mariana Mazzucato, The Value of Everything, Allen Lane, 2018. Mary Poovey, A History of the Modern Fact, University of Chicago Press, 1998. Staffan Andersson and Paul M. Heywood, "The Politics of Perception," Political Studies 57(4), 2009, 746-767. Transparency International CPI 2024 methodology note, accessible via transparency.org. UNODC SDG indicators 16.5.1 and 16.5.2 metadata. OECD Better Life Index launched May 2011. New Zealand Treasury, Wellbeing Budget 2019. Companion to The Office Is the Dose, A Poor Politician with Excellent Media Training, Both Halves of the Apparatus and The Standards Were Never Meant to Be Met. FSI exact secrecy score for Ireland flagged for re-verification before next release cycle.

Overwatch Report is an independent publication. We have no financial positions in any entity mentioned.