Why this piece exists

The most common defect in Irish political conversation is impatience with the wrong things. We become impatient with policies that have been in operation for two years (it is too early to tell), and patient with policies that have been in operation for forty (it is too late to change). We treat as urgent the things that move on a five-year electoral cycle, and as eternal the things that take a generation to shift. We ask "why has nothing changed?" about questions where the change has been continuous and substantial, and we ask "why is everyone so calm about this?" about questions where the actual time-horizon for change is decades.

This piece argues that the most useful single calibration tool for Irish political thinking is the 25-year rule. Major systemic transitions, from one entrenched arrangement to its replacement, tend to take roughly twenty to twenty-five years from the moment the old arrangement begins serious decline to the moment the new arrangement is fully consolidated. The rule is not a law. It is a heuristic with substantial historical support. Used carefully, it prevents both despair (it is possible) and impatience (it takes longer than one electoral cycle).

The piece does three things. It explains the structural reasons the rule holds. It walks through the historical evidence. And it applies the rule to four current Irish strategic questions: unification, the climate-and-energy transition, AI literacy, and the housing-and-property settlement. The application is not a forecast. It is a calibration: if these are 25-year transitions, what does that mean for what we should be doing right now?

Why the rule holds

The 25-year figure is not coincidence. It tracks three independent structural features of how human societies actually change.

Generational turnover. A generation is roughly 20 to 25 years. Entrenched systems are upheld by incumbents whose careers, networks, mental models, and identities are bound to those systems. They do not abandon the systems. They retire from them. Until a full new generation, educated and socialised under the new paradigm, has come of age and assumed institutional positions, the old arrangements remain operationally dominant. This is true even when the formal ideology has changed. The generation that grew up under the old system continues to run the institutions in the old system's image.

The S-curve of adoption. Everett Rogers's Diffusion of Innovations (1962, expanded across subsequent editions) decomposes the adoption of any new system into three phases: innovators and early adopters (five to ten years), early and late majority (ten to fifteen years), and laggards (five to ten years). For socioeconomic systems at scale, the full S-curve runs roughly 20 to 30 years. Faster than this is the exception (digital products with low switching costs, where ten years is plausible). Slower than this is also the exception (capital-intensive infrastructure with long amortisation, where forty years is plausible).

Capital amortisation. Major systemic infrastructure (railways, factories, power stations, building stock, institutional architecture) is built on financing assumptions that depreciate over 25 to 50 years. The capital does not become available for redirection until those cycles complete. New systems either build alongside the old until the old amortises, or wait until the old fails and resources free up. Both pathways take time.

These three features compound. The generational and adoption curves run in parallel; the capital cycle either reinforces or constrains them. The 25-year window is the band where all three converge for any transition that is genuinely systemic rather than incremental.

The historical evidence

The rule is well-supported across modern history. A representative sample.

Railways replacing canals (1830s to 1860s). First major trunk lines in Britain in the 1830s. By the 1860s most freight and passenger traffic had shifted from waterways to rails. Roughly thirty years from take-off to consolidation. The canal system did not disappear; it was simply no longer the primary mode.

Automobiles displacing horses (1900 to 1925). First mass-production cars in the 1900s. By the mid-1920s, urban transport was dominated by motor vehicles, and horse-drawn traffic was a marginal presence. Twenty-five years.

Meiji Japan (1868 to 1895). Tokugawa shogunate dismantled in 1868. Modern industrial economy, conscript army, universal primary education, constitutional government, and rail network in place by the mid-1890s. Twenty-five years.

Post-war Western Europe (1945 to 1970). From war's end to the maturation of the social-democratic settlement, the welfare state, the European Common Market, and the consumer economy. Twenty-five years.

China's Reform Era (1978 to 2003). Deng's reforms launched in 1978. By the early 2000s, the Maoist commune system was substantially dismantled, market pricing was operational across most of the economy, special economic zones had matured, and China was a major player in the global industrial economy. Twenty-five years.

South Korea's industrialisation (1960 to 1985). From agrarian post-war recovery to chaebol-led heavy industry, export-oriented growth, and full urbanisation. Twenty-five years.

Eastern Europe after 1989 (1989 to 2015). Berlin Wall in 1989. EU and NATO consolidation across most of the former Eastern Bloc by the mid-2010s. Twenty-five years for full political-economic integration.

Mobile telephony in sub-Saharan Africa (2000 to 2025). From sparse state-owned landlines covering under 5 percent of population to near-universal mobile penetration with mobile-money infrastructure. Twenty-five years.

The Irish-specific cases are particularly worth listing because they are the cases the rule is being applied to in the second half of this piece.

The Land League and the Land Acts (1879 to 1903). From the start of mass tenant mobilisation to the Wyndham Act that completed the transfer of most Irish farmland from landlord to tenant ownership. Twenty-four years.

The slow liberalisation of Irish social life (1973 to 1995). Contraception ban first contested in the McGee case in 1973; substantial liberalisation through the 1980s; divorce permitted by referendum in 1995. Twenty-two years for the central change.

The professionalisation of the Irish economy (1958 to 1985). From Whitaker's Economic Development and Lemass's outward turn through the IDA build-out, EEC accession in 1973, and the consolidation of the FDI-led economic model by the mid-1980s. Twenty-seven years.

The unwinding of clerical political authority (1990 to 2015). Beginning with the early-1990s scandals and the slow exposure of institutional abuse, through the 2003 Ferns Report, the 2009 Murphy Report, and the 2015 marriage equality referendum. Twenty-five years for the shift from clerical authority to secular liberalism as the dominant Irish public framing.

The pattern holds across cases that are otherwise quite different in scale, region, and content. Twenty-five years is the unit of measurement for systemic transition.

The caveats

The rule is a heuristic, not a law. Three caveats matter.

Scale matters. Incremental product replacements (VHS to DVD, taxi to ride-share, retail to e-commerce) can happen in ten to fifteen years. Fundamental systemic shifts (fossil fuels to renewables, industrial agriculture to alternative protein) may take thirty to fifty years because of infrastructure lock-in, regulation, and capital intensity. The 25-year figure is the central tendency; the variance is real.

Ecosystem resistance. Entrenched systems sometimes mount sustained legal, political, and cultural defences that lengthen the cycle. The fossil-fuel industry's well-documented post-1970s campaign to delay climate response is the canonical example. The Irish corporate-tax model's three-decade defence against international tax-coordination pressure is a smaller-scale local example. Where the incumbent system is wealthy and politically organised, the cycle stretches.

Non-linear acceleration. Some transitions, particularly digital ones with low switching costs and network effects, compress the curve. Smartphones went from niche to near-universal in under ten years. Generative AI has, at the time of writing in 2026, achieved substantial mainstream penetration in under five years from the public launch of ChatGPT in November 2022. The 25-year frame is the default; specific cases can run faster when the underlying technology and economics permit.

The honest application of the rule is therefore: 25 years as the starting hypothesis, with explicit consideration of whether the case at hand is incremental (faster), capital-intensive (slower), politically-defended (slower), or digitally-accelerated (faster). The starting hypothesis is the contribution. Refining it is the user's work.

The four current Irish strategic transitions

Now to the application. Four major Irish strategic questions are currently in some phase of transition. The rule lets us calibrate where each one actually is, what the realistic horizon for completion is, and what the implication is for what we should be doing now.

Irish unification

The transition. From the partition settlement of 1922 to whatever post-partition arrangement Ireland arrives at, whether that is full unification, federation, an enduring devolved Northern Ireland inside the UK, or some currently-unimagined alternative.

Where the curve sits. The 1998 Belfast Agreement is a plausible take-off point for the modern unification process. The Agreement provides the procedural mechanism (a border poll), defines the conditions, and embeds the principle of consent. From 1998 to 2023, the demographic, political, and constitutional conditions for a serious unification conversation have been accumulating. The 2023 Northern Assembly elections, the 2024 demographic projections, the post-Brexit legal complications, and the ongoing constitutional review work in the South all sit inside this curve.

Where the rule places us. If 1998 is the take-off, the rule places consolidation around 2023 to 2028. The current period is precisely when, on the rule, the substantive transition would be expected to either occur or to be visibly proximate. This is consistent with the empirical record: serious unification conversation has moved from background to foreground over the last three years, the political demand for a border poll within the next decade is now mainstream rather than fringe, and the institutional preparation work (the Constitutional Future of the Island consultations, the various Senator Mark Daly Seanad reports, the academic work by Brendan O'Leary and others) is substantial.

What this implies. The unification question is not a long-term question. It is a current-period question. The rule suggests the substantive decisions about whether and how unification occurs will be taken in the late 2020s and consolidated in the early-to-mid 2030s. This is not 50 or 100 years away. It is one-to-two electoral cycles away. The political-intellectual work of drafting the 32-county republic's constitutional, fiscal, and social architecture is the work of now, not the work of later.

The calibration check. Is unification capital-intensive in a way that would slow the curve? Substantially. The fiscal transfer architecture, the institutional integration, the public-sector reorganisation are all capital-heavy. Is it politically defended? The unionist position is, by definition, structurally opposed to unification, and the British state has its own interests in the existing arrangement. These two factors push the rule's expected timing later rather than earlier. The realistic window is therefore plausibly the early-to-mid 2030s for the decision and the late 2030s for full institutional integration. Either way, this is a question for the current generation, not a question for a hypothetical future.

The climate-and-energy transition

The transition. From a fossil-fuel-based energy and transport system, with carbon-intensive agriculture and built environment, to a low-carbon equivalent.

Where the curve sits. Variable depending on which sub-system one tracks. Electricity generation is well into the curve in Ireland (wind generation has grown substantially over the last fifteen years; the grid is in mid-transition). Domestic heating and transport are at earlier phases. Agriculture is at a very early phase. Each sub-system has its own curve.

Where the rule places us. This is one of the cases where the rule's "fundamental systemic shift" caveat applies and the curve is plausibly 30 to 50 years rather than 25. The infrastructure lock-in is severe. Ireland's gas network was built across the late 20th century and is amortised over a 50-year horizon. The vehicle fleet turns over in 15 to 20 years. The built environment turns over in 50 to 100 years. The agricultural sector's structural transformation is generational. The Irish-specific deadlines (Climate Action Plan 2030 targets, EU Fit-for-55 commitments, the 2050 net-zero goal) are not aligned with the natural curve and will require either accelerated capital amortisation (premature retirement of fossil infrastructure) or missed targets.

What this implies. The climate transition is the strategic question where the rule's caution is most needed. It does not mean the transition is impossible. It means that the transition's natural pace is slower than the policy timetable, and that closing the gap requires either accelerated capital write-offs (politically expensive), accelerated regulatory pressure (subject to capture), or accelerated public investment (subject to fiscal constraint). The Irish political conversation about climate policy mostly does not engage with this honestly. The conversation is conducted in terms of either underestimating the difficulty or overstating the inevitability of failure. The rule suggests neither is correct: the transition is achievable on a 30 to 50 year timescale, with substantial capital and institutional cost.

AI literacy and the AI-economic transition

The transition. From an information-and-knowledge economy where most cognitive work is done by humans, to a hybrid economy where AI systems perform substantial fractions of the cognitive work and humans perform supervisory, creative, and relational roles. This includes the related question of public AI literacy: whether the population can engage with this technology as adults rather than as subjects.

Where the curve sits. Take-off is the November 2022 ChatGPT public launch. By any reasonable count, we are roughly 3.5 years into the curve at the time of writing in May 2026. The rule's standard timing would place full consolidation around 2042 to 2047. The digital-acceleration caveat may compress this to 2032 to 2037. Either way, the transition is in early phase, with most of the curve still ahead.

Where the rule places us. This is one of the cases where digital acceleration is real. The supply side has matured at a pace that the standard 25-year curve substantially underestimates. The deployment side is moving faster than any prior major-technology adoption curve in modern history. The labour-market and educational implications are arriving over years rather than decades.

What this implies. The implications for Ireland specifically are large and largely unprepared-for. The Irish economy is unusually exposed to AI because of the FDI-tech-platform model: a substantial fraction of the European AI ecosystem operates on Irish soil. The Irish education system's preparation for this transition is, at the time of writing, minimal. Computer science teaching at second level is patchy and framework-mediated rather than first-principles-based. The adult-education provision in AI literacy is essentially absent.

The rule, accelerated, suggests that the consolidation phase of the AI transition will arrive in the early-to-mid 2030s. The institutional work of preparing for it is the work of the current period. This includes both the labour-market transition (which jobs change, which disappear, which emerge) and the public-literacy transition (whether the population can participate in decisions about AI policy as informed citizens or only as objects of policy made elsewhere).

The calibration check. The acceleration is real but not unlimited. The deployment of AI is faster than the deployment of regulation and infrastructure to support it. The institutional response will run on the standard timescale even where the technology runs on the accelerated one. The realistic frame is therefore: the technology consolidates by the mid-2030s; the institutional response consolidates by the mid-2040s. The gap between these two is the period of maximum disruption and the period in which intelligent intervention has the most leverage.

The Irish housing-and-property settlement

The transition. From the current arrangement (housing as primary investment-asset class, concentrated ownership, planning-system stasis, supply chronically below demand) to whatever replacement arrangement Ireland arrives at. The replacement could be Nordic-mode universal public-and-cooperative housing, distributist-mode wide-distributed-ownership, neoliberal-mode pure market clearing, or some hybrid.

Where the curve sits. This is the strategic question where the take-off point is hardest to identify. Several plausible candidates: 1996 (the start of the Celtic Tiger property inflation), 2008 (the crash and bank guarantee that locked in the current asset structure), 2020 (the post-Covid acute-crisis acceleration), or 2024 (the regulatory openings, including the backyard-prefab change). Different take-off dates produce different rule-based estimates.

The most plausible reading. The current housing crisis emerged from the post-2008 settlement (the bank guarantee, the NAMA-era property reorganisation, the post-2014 institutional landlord acquisition wave). Take-off as a coherent crisis is roughly 2014. The rule places consolidation of the response around 2034 to 2039. Substantive policy change will occur within this window.

What this implies. The rule suggests the housing crisis is not a permanent feature. It also suggests that the current period (mid-2020s) is mid-curve rather than late-curve. The housing settlement that replaces the current one is being shaped now, in the form of the 2024 prefab regulation, the cuckoo-fund tax changes, the developing public-housing build-out, the cooperative-housing experiments at small scale, and the broader political-cultural conversation about whether housing is an asset class or a public good. The shape of the consolidated outcome depends on which of these strands prevails over the next decade.

The calibration check. Is housing a capital-intensive transition? Yes, severely. Each unit of housing is a long-term capital asset. The existing housing stock is owned by people and institutions whose economic interests are tied to its continued treatment as an asset class. Is it politically defended? Yes; the asset-owning class includes most of the Irish electorate by direct or indirect interest. These two factors lengthen the curve. The realistic timing for full settlement consolidation is plausibly the early 2040s rather than the mid-2030s.

The substantive choices, however, are being made now. Whether Ireland builds toward Nordic-public, distributist-cooperative, or neoliberal-market consolidation is determined by decisions taken in the current decade. The consolidation is later. The trajectory is now.

What this means for current political work

The rule has practical implications for how Irish citizens and Irish political institutions should be operating in the current period. Three implications stand out.

The work of the current period is preparation, not deferral. All four major strategic transitions have their consolidation phases in the 2030s and 2040s. The decisions that shape those consolidations are being made in the late 2020s. This means the strategic work is now. The policy proposals, the institutional designs, the political coalitions, and the public-literacy infrastructure that will shape the consolidation phase have to be built in the current period or they will not be available when needed.

This argues against two common Irish political postures. The first is "we have time, we will deal with it later". On the rule, we do not have time in this sense; the determining decisions are immediate. The second is "the situation is hopeless because nothing has changed in five years". On the rule, five years is too short an interval to expect systemic change to be visible. The work is happening on a 25-year horizon and judging it on a 5-year horizon misreads the timescale.

The four transitions are entangled. The unification question depends on the housing question (any unified Ireland inherits the Northern social-housing tradition and the Southern asset-housing arrangement, which are structurally incompatible). The climate transition depends on the housing question (the residential built environment is roughly a third of the carbon problem). The AI transition depends on all of them (labour-market disruption affects housing affordability, climate-policy implementation affects energy systems that AI infrastructure depends on, unification affects which institutions are doing the policy work). Treating the four as separate problems with separate timelines underestimates their interaction. The rule suggests they will consolidate roughly together, in the 2030s and 2040s, and the institutional architecture that handles one will substantially shape what is possible for the others.

The political-literacy work is the long-cycle work. Major systemic transitions do not consolidate around policy alone. They consolidate around shared frameworks for thinking, public vocabularies for naming what is happening, institutional cultures for handling new conditions, and educational systems that produce citizens able to participate. The Political Literacy programme on this site is, in part, an attempt to do this kind of long-cycle work for the Irish context. The 25-year horizon makes the value of such work legible: even where individual articles do not produce immediate policy impact, the cumulative effect over a decade is the substrate that makes the consolidation phase possible.

This is what the Distributism piece, the Nationalism piece, the Republicanism piece, and the others in the Political Literacy series are for. They are not pamphlets for the next election. They are vocabulary for the next two decades.

The honest assessment

The 25-year rule is a heuristic. Specific cases will run faster or slower, and the rule does not absolve anyone from the work of looking at the specific factors that affect a given transition.

The rule's main contribution is calibration. Used carefully, it prevents the two most damaging postures in modern political conversation: the impatience that judges decade-scale transitions on year-scale evidence, and the despair that concludes nothing can change because nothing has changed quickly. Both are misreadings of how human societies actually transform. The history says transformation is possible and takes a generation. The rule encodes that history into a usable working tool.

For Ireland specifically, the rule places all four major strategic questions inside a roughly 2030 to 2045 consolidation window. The work of the current period is the work that determines what consolidates. This is not a long-term question. It is the immediate work of the next two electoral cycles, conducted on a horizon that recognises what those electoral cycles are actually for.

A reader who has worked through this piece will leave with a calmer and more accurate sense of the time-horizon for Irish political change than is available in most current public-discourse sources. That is the contribution. The framework is not original to me; it is drawn from Rogers, from generational-turnover demography, from capital-amortisation economics, and from a long tradition of comparative-historical analysis. The contribution is the application to current Irish strategic questions, in a form that lets readers calibrate their own expectations.