{ "title": "Three versions of the good life", "subtitle": "More prosperity. A longer life. More room to choose.", "standfirst": "The United States, Canada and the European Union change places when the question changes. Thirty-nine exhibits and three linked instruments compare prosperity, institutions, health, learning, defence and AI. Follow the Rank Weave, inspect the Relationship Atlas, then test where Canada fits at the Atlantic Seam.", "date": "2026-09-17", "revision": "atlantic-r6", "readTime": "45 min", "executive": { "title": "Three societies, in a nutshell.", "deck": "The United States has the largest resource base per person. Canada combines high educational attainment with unusually concentrated export exposure. The EU pairs longer lives with substantial differences among its members. Eight common KPIs, four miniature comparisons and the trade links put these different positions in view; the chapters explain what each measure leaves out.", "countries": [ { "country": "CAN", "kicker": "The Canadian bulletin", "title": "A smaller economy. A concentrated relationship.", "interpretation": "Canada's output per person is above the EU aggregate, but its recent per-person recovery is weaker. High tertiary attainment coexists with an export book dominated by the United States.", "anchor": "dependence", "takeaway": "Watch the denominator and the destination." }, { "country": "USA", "kicker": "The American bulletin", "title": "The resource leader. An outcome puzzle.", "interpretation": "The United States produces more per person and spends more on health. The matched health comparison nevertheless records shorter lives. Resources alone do not settle the institutional question.", "anchor": "health", "takeaway": "Distinguish capacity from conversion." }, { "country": "EUU", "kicker": "The European bulletin", "title": "One union. Twenty-seven versions.", "interpretation": "The EU aggregate records longer lives at lower health spending than the United States. Its average conceals substantial national variation and describes neither a single treasury nor one school system.", "anchor": "resources", "takeaway": "Open the aggregate before drawing a verdict." } ] }, "panels": [ { "id": "resources-denominator", "chapter": "resources", "afterGroup": 1, "country": "CAN", "label": "The denominator matters", "title": "An economy can expand while output per person falls.", "body": "Total production and population can move in the same direction at different speeds. Canada's 2024 contrast makes that accounting visible. It does not identify the causal effect of immigration; investment, participation and productivity would also change under another population path.", "fact": "2024: total GDP +2.05%; per person -0.95%", "sourceFigure": 5 }, { "id": "resources-distribution", "chapter": "resources", "afterGroup": 2, "country": "EUU", "label": "Inside the European average", "title": "The union is not a representative country.", "body": "A country-level distribution gives each member one observation, regardless of its population. That is useful for seeing heterogeneity, but it is not a distribution of household incomes. Luxembourg and Ireland also remind us that territorial output need not describe residents' material circumstances particularly well.", "fact": "Country dispersion is not household dispersion.", "sourceFigure": 6 }, { "id": "public-overlap", "chapter": "public-choices", "afterGroup": 0, "country": "USA", "label": "Inputs into capability", "title": "More investment is a beginning, not a result.", "body": "A spending share tells us how many resources were committed, not what those resources produced. The assets, their use, complementary skills and organizational change matter. Research expenditure is similarly an input into innovation rather than a measure of successful innovation.", "fact": "US R&D: 3.45% of GDP in 2023", "sourceFigure": 8 }, { "id": "public-treasuries", "chapter": "public-choices", "afterGroup": 1, "country": "EUU", "label": "Twenty-seven treasuries", "title": "This is not the budget in Brussels.", "body": "The EU fiscal figure weights national-government expenditure ratios by each member's nominal GDP. It is not the budget of EU institutions. Public consumption, defense, health, R&D and investment also overlap: adding these percentages would manufacture a national spending pie that the definitions do not support.", "fact": "EU national-government expenditure: 49.02% of GDP / 2024", "sourceFigure": 11 }, { "id": "health-conversion", "chapter": "health", "afterGroup": 0, "country": "USA", "label": "The conversion question", "title": "Higher spending. A shorter measured life.", "body": "The United States spends 2.32 times the EU amount per person on health in the matched 2023 snapshot, while life expectancy is about three years lower. Prices, access, disease burdens, inequality and risks outside health care can all matter. The plotted paths cannot assign causal credit to any one explanation.", "fact": "A descriptive reversal, not an efficiency frontier.", "sourceFigure": 12 }, { "id": "health-context", "chapter": "health", "afterGroup": 2, "country": "EUU", "label": "A wider comparison", "title": "Keep the major comparison. Expose some of its variation.", "body": "Germany, France, Sweden, Poland and Spain appear as subdued contextual trajectories. They come from the same frozen World Bank extracts and are already included in the EU aggregate. They are a transparent subset for comparison, not five independent replications of a European effect.", "fact": "The gray paths are context, not a fitted control group.", "sourceFigure": 15 }, { "id": "learning-profile", "chapter": "learning", "afterGroup": 0, "country": "EUU", "label": "One system, three subjects", "title": "Follow a profile, not an invented EU score.", "body": "The subject-profile graphic connects science, reading and mathematics on the same score scale. Those links compare different competencies, not years or a student's progress. Every EU member remains a country observation; hovering or keyboard inspection brings one national profile into focus.", "fact": "Three subject axes. No pooled EU mean.", "sourceFigure": 19 }, { "id": "learning-tails", "chapter": "learning", "afterGroup": 1, "country": "CAN", "label": "The average is not the distribution", "title": "The same reading mean can leave different tails.", "body": "Canada and the United States both report a reading mean of 490. Their reported P90-P10 spreads are 273 and 312 points. The median in the percentile graphic is a different statistic, and sampling cautions remain attached. The plot cannot identify how much of the spread comes from family background or school organization.", "fact": "490 points in both countries; different score distributions.", "sourceFigure": 20 }, { "id": "constraints-age", "chapter": "constraints", "afterGroup": 0, "country": "EUU", "label": "A population composition", "title": "An older share is not a forecast of decline.", "body": "Each square in the age composition stands for one percentage point, not one observed person. The colored share is age 65 and over; the remainder is under 65. The older comparison date remains visible, but participation, migration, health and productivity still shape what aging will mean.", "fact": "Composition describes the starting point, not the outcome.", "sourceFigure": 23 }, { "id": "constraints-carbon", "chapter": "constraints", "afterGroup": 1, "country": "CAN", "label": "The emissions boundary", "title": "Where production happens changes the account.", "body": "Territorial CO2 belongs to the place where emissions occur. A consumption account also follows emissions embodied in imports and exports. Falling domestic emissions can be meaningful without being a complete account of everything residents consume.", "fact": "Territorial CO2, excluding land-use change.", "sourceFigure": 26 }, { "id": "dependence-asymmetry", "chapter": "dependence", "afterGroup": 0, "country": "CAN", "label": "The asymmetric relationship", "title": "The same customer can matter very differently to each side.", "body": "The United States takes 72.49% of Canada's merchandise exports. Canada takes 15.40% of US exports. The chord makes the connections visible; the export-share view asks how dependent each economy is on them. Those are different denominators, not competing versions of one statistic.", "fact": "Canada to US: 72.49% of the Canadian export book / 2025", "sourceFigure": 28 }, { "id": "dependence-reporters", "chapter": "dependence", "afterGroup": 1, "country": "USA", "label": "Two reporting perspectives", "title": "A mirror gap is a reason to inspect the definition.", "body": "Exporter and importer reports differ through valuation, partner attribution, transit, timing and revisions. The story retains the exporter perspective in the chord and the alternative reports in the final exhibit. It never averages the two into a figure that no agency actually reported.", "fact": "Reporter differences are not a trade balance or missing goods.", "sourceFigure": 29 } ], "opening": [ "A Canadian traveler can change a destination much faster than a Canadian factory can change its customers. The earlier aviation story followed a recovery in flying that was changing direction. This comparison asks a larger question: how different are the societies on either side of the Atlantic, and how much freedom do they have to choose another course?", "The clearest result appears before we reach any judgment about their institutions. In 2024, US output per person was about 39% above the EU aggregate and 32% above Canada after adjusting for purchasing power. In the matched 2023 health comparison, life expectancy was roughly three years lower in the United States than in either comparator, despite substantially higher health spending per person. Economic resources and lived outcomes describe different orderings.", "That reversal is a reason to investigate how prosperity reaches people. It is not a causal estimate of what a government, a health system or a culture contributes. This story keeps the measures separate, carries the same three reference points through the evidence and opens the European aggregate wherever it conceals the important comparison.", "There is no overall welfare score. Producing one would require a judgment about how much output compensates for a shorter life, how credentials compare with learning, and how much dependence a society should accept. The extension asks a different question: how similar are the measured profiles? Its weights, normalization and sensitivity remain explicit rather than becoming an identity verdict." ], "chapters": [ { "id": "resources", "number": "01", "label": "Resources", "title": "A large economy is not the same as a prosperous person.", "deck": "First establish the scale. Then change the denominator.", "paragraphs": [ "The European Union has the largest population of the three, the United States the largest economy at market exchange rates, and Canada a much smaller domestic market. These facts matter for procurement, bargaining power, the fixed costs of research and the reach of a national firm. They do not describe the resources available to a typical household.", "The EU here always means its current 27 members, excluding the United Kingdom. That statistical perimeter is deliberately fixed. A union of countries is nevertheless not a federation with one treasury, one labor market or one public-service system. Comparable columns are a device for asking questions, not a claim that their institutions are equivalent." ], "groups": [ { "figures": [ 1, 2 ], "layout": "sequence", "paragraphs": [ "The purchasing-power comparison changes both the units and the question. In constant 2021 international dollars, output per person in 2024 was approximately $75,698 in the United States, $57,534 in Canada and $54,342 in the EU. These are common purchasing-power units, not dollars that can be exchanged at a bank. The series adjusts for price differences across places and inflation through time.", "GDP measures production within an economy. It is not median disposable income, household wealth, the value of unpaid care or leisure. An economy can produce more because workers produce more per hour, because more people work, because they work longer, or through some combination of those margins. The American advantage is real in this measure; the measure cannot tell us who receives it." ] }, { "figures": [ 3, 4 ], "layout": "sequence", "paragraphs": [ "Levels and growth also answer different questions. Between 2000 and 2024, real output per person rose by about 37.5% in the United States, 35.3% in the EU and 18.9% in Canada. Similar proportional progress can leave a substantial initial level gap in place. Conversely, a high starting level can coexist with a disappointing recent path.", "Canada's 2024 output per person was slightly below its 2019 level even though the economy as a whole had expanded. Population belongs in the accounting. This observation does not identify the effect of immigration: it does not tell us what output, investment, participation or productivity would have been under another migration path. Treating an accounting denominator as a causal explanation would skip the central empirical question." ] }, { "figures": [ 5, 6 ], "layout": "sequence", "paragraphs": [ "The European aggregate needs a second qualification. Member-country output per person ranges from about $34,221 in Bulgaria to $128,475 in Luxembourg. Ireland also lies far above the aggregate. Multinational production and cross-border labor can make territorial output an especially imperfect description of residents' material circumstances.", "The distribution is therefore a distribution of countries, with one observation per country. It is not a distribution of European households. The distinction matters: an equal-country view gives a small state the same visual presence as a large one, whereas the aggregate answers a population-level production question." ] } ], "takeaway": "Keep three questions separate: how large is the economy, how much does it produce per person, and how has that changed?" }, { "id": "public-choices", "number": "02", "label": "Public choices", "title": "Resources create possibilities. Their allocation matters.", "deck": "Investment, research, public services and defense are different commitments, not pieces of one additive pie.", "paragraphs": [ "Canada devoted a larger share of GDP to fixed investment in 2024 than either the United States or the EU, yet it did not record the strongest recent growth in output per person. A higher investment share is not a sufficient explanation of future prosperity. The next questions concern the assets being created, their utilization, their returns and the capabilities that complement them.", "A national investment aggregate includes buildings, equipment and other fixed assets across public and private sectors. Its composition is consequential, but this package does not decompose that composition finely enough to allocate responsibility. The chart establishes the question rather than supplying a housing, technology or infrastructure explanation by assumption." ], "groups": [ { "figures": [ 7, 8 ], "layout": "pair", "paragraphs": [ "Research and development expenditure gives a more distinct ordering. In the common 2023 observation, the United States devoted 3.45% of GDP to R&D, against 2.26% in the EU and 1.81% in Canada. The rank path shows where the ordering changed; the percentages show how large the differences are. A rank alone cannot do both jobs.", "R&D spending is an input into discovery and development, not a count of successful innovations. The benefits can accrue abroad, take years to appear or depend on complementary skills and organizational change. Still, the expenditure difference describes a material choice about how much of today's output is committed to future capabilities." ] }, { "figures": [ 9, 10, 11 ], "layout": "three", "paragraphs": [ "Public-service consumption and total government expenditure must not be treated as synonyms. Government final consumption concerns services produced or purchased for current use. General government expenditure also includes transfers and other uses. The latter reached approximately 49.0% of GDP for the EU comparison, 44.3% in Canada and 37.7% in the United States in 2024.", "The EU expenditure ratio is a GDP-weighted calculation across 27 national governments using IMF data. It is not the budget of EU institutions. Neither a high nor a low ratio, by itself, establishes the quality of public services, the distribution of support or the efficiency with which programs operate.", "Defense starts from a different allocation again. The historical 2024 military-expenditure shares were 3.42% of GDP in the United States, 1.90% for the EU aggregate and 1.31% in Canada. These are observed expenditure concepts, not measures of readiness, NATO-compliant accounting or forecasts of later commitments. An announcement belongs to a different evidence category.", "These measures overlap. Government spending can finance health, education, military activity and investment; R&D can overlap with investment. Adding them into a national spending pie would create double counting precisely where the comparison needs conceptual clarity." ] } ], "takeaway": "A spending share identifies a commitment of resources. It does not, without another layer of evidence, identify what that commitment achieved." }, { "id": "health", "number": "03", "label": "A longer life", "title": "The richest economy does not deliver the longest lives.", "deck": "The strongest reversal in the comparison is also where causal restraint matters most.", "paragraphs": [ "In 2023, current health expenditure per person was approximately 13,473 international dollars in the United States, 7,355 in Canada and 5,806 in the EU. Life expectancy in the same year was 78.4, 81.6 and 81.4 years respectively. Matching the year removes one avoidable source of disagreement. It does not make the remaining explanation simple.", "The health dollar is a current international dollar at purchasing power parity. The earlier GDP-per-person series uses constant 2021 international dollars. Dividing one by the other would combine incompatible price bases. Likewise, the health-spending trajectory contains price changes and must not be described as real expenditure growth." ], "groups": [ { "figures": [ 12 ], "layout": "feature", "paragraphs": [ "The connected observations show the path each society followed; they are not an estimated production frontier. A point further right represents higher measured expenditure. A point further up represents longer period life expectancy. Neither position identifies how much health an additional dollar would buy.", "Health outcomes reflect health care alongside income distribution, living conditions, disease burdens, behavior, violence, transport risks and the history of public health. Higher spending can also reflect higher prices. Three aggregate trajectories cannot separate these mechanisms. The defensible result is narrower and important: more output and more health spending do not guarantee longer lives." ] }, { "figures": [ 13, 14 ], "layout": "pair", "paragraphs": [ "Life expectancy at birth is a period statistic. It summarizes how long a hypothetical person would live under the mortality rates measured in that period; it is not a forecast of the life span that a particular newborn will realize. Short-run shocks and their reversal therefore belong in the interpretation.", "The difference between women and men adds another level of variation within every comparator. A national average can conceal a substantial gap even before income, place or other social conditions enter the analysis. The figure describes the sex-specific series available in the source. It is not a causal estimate of any one risk factor." ] }, { "figures": [ 15, 16 ], "layout": "pair", "paragraphs": [ "Child mortality asks a different, more bounded question. The probability of dying before age five fell across all three comparators, but the gap remains. The 2024 estimates were 6.5 per 1,000 live births in the United States, 5.4 in Canada and 3.86 in the EU. These are official model-based mortality probabilities, not a direct count of deaths in a single birth cohort.", "Financing arrangements provide context without resolving the outcome comparison. Government financing, out-of-pocket payments and the residual category sum to current health expenditure. The residual must not simply be renamed private insurance. Nor is an out-of-pocket share a measure of unmet need or the financial burden on a low-income household: its denominator is total expenditure, not household income." ] }, { "figures": [ 17 ], "layout": "feature", "paragraphs": [ "Europe's position should not become a reassuring curtain. The aggregate sits among member countries with different mortality histories and living conditions. Showing those countries prevents a continental mean from standing in for every resident, while the national scope of the observations prevents us from claiming to have mapped inequality within each country.", "A serious health-system comparison would next examine prices, access, preventable and treatable mortality, age structure and the distribution of outcomes. The present evidence is enough to reject the idea that a resource ranking settles the question. It is not enough to award causal credit." ] } ], "takeaway": "The spending-outcome reversal is an empirical finding. A league table of health-system efficiency would be a different, unestimated claim." }, { "id": "learning", "number": "04", "label": "The next generation", "title": "A qualification, a test score and an opportunity are different things.", "deck": "Education changes the comparison again. It also makes the limits of aggregation unusually visible.", "paragraphs": [ "Education spending records resources committed to learning. Tertiary attainment records completed qualifications. PISA measures selected skills among tested 15-year-olds. These measures concern different populations, periods and outcomes; one cannot serve as a direct validation of another.", "Even the spending comparison has an unusual European denominator. The 2021 WDI EU value is the median of 26 reporting member-country expenditure shares. It is explicitly a member-country median, not the ratio of all EU education expenditure to all EU output. Canada and the United States remain national ratios." ], "groups": [ { "figures": [ 18, 19 ], "layout": "sequence", "paragraphs": [ "The PISA exhibits retain the published 2025 estimates from the official CMEC national report, including its international comparison tables. An independent check matched the official report's edition and numerical tables to the frozen extracts. This is not a new analysis of student records. Published standard errors, confidence limits and source sampling cautions travel with the estimates.", "Sampling cautions matter especially for Canada, the United States and the Netherlands. Canada's weighted student response rate was 77.4%, below the 80% standard, and the report identifies potential upward student-level bias. The other two systems retain their source flags; their precise adjudication wording was not independently re-extracted. An unflagged row is not a certificate of perfect comparability.", "There is no manufactured EU student mean. An unweighted mean of country scores would answer a different question from a student-weighted estimate, and neither can be substituted silently for the source's country observations. The table shows the European systems themselves alongside Canada and the United States.", "The subject matters. Science, reading and mathematics measure related but distinct competencies. The connected subject profiles make those differences visible on one common score scale, with published uncertainty retained. The connecting lines compare subjects, not time or student progress; no pooled European score is constructed." ] }, { "figures": [ 20, 21 ], "layout": "sequence", "paragraphs": [ "Canada and the United States both had a reported reading mean of 490 points. Yet the reported distance between the 90th and 10th percentiles was 273 points in Canada and 312 in the United States. Similar averages can accompany different spreads. In the percentile figure, the central mark is the median, not that mean.", "The published percentile gap uses underlying values before rounding. Subtracting the displayed rounded endpoints can therefore differ by a point from the reported gap. Both belong in the exact-value record, with their roles kept separate. The central box runs from the 25th to the 75th percentile; the outer interval runs from the 10th to the 90th. Neither is a confidence interval.", "The mathematics comparison uses another kind of interval: published uncertainty around a country mean. It does not show the spread of individual students' scores. Pointwise confidence intervals also do not establish simultaneous significance for every pair of countries in a long ranking.", "Nothing in these distributions identifies the contribution of family background, school organization or a particular policy. Estimating a new socioeconomic gradient would require student records, plausible values, replicate weights and a suitable design. The education escalator story examines the opportunity question more directly; this comparison establishes why it cannot be settled with the average alone." ] }, { "figures": [ 22 ], "layout": "feature", "paragraphs": [ "Credentials produce another ordering. Tertiary attainment among people aged 25-34 was 69% in Canada, 52.7% in the United States and 43.9% in the EU in 2024. These figures include short-cycle qualifications as well as degrees. National survey populations and coverage are not perfectly identical, and Canada's published whole-percentage precision should remain visible.", "The people in this attainment comparison are not the 15-year-olds in the PISA assessment. Migration, completion, cohort replacement and the structure of qualifications can all affect attainment. A high credential share does not replicate a test-score comparison, just as a score distribution cannot be read as a distribution of adult earnings." ] } ], "takeaway": "Keep the input, the credential, the demonstrated skill and the distribution of opportunity in separate analytical columns." }, { "id": "constraints", "number": "05", "label": "Future constraints", "title": "The next decade starts from different constraints.", "deck": "Age structure, fertility and territorial emissions qualify the national success stories without determining their futures.", "paragraphs": [ "People aged 65 and over represented approximately 22.0% of the EU population, 19.8% of Canada's and 17.9% of the US population in 2024. These age structures matter when interpreting employment rates, health demand and fiscal pressures. A comparison that ignores them can attribute a compositional difference to institutional performance.", "Age structure is not destiny. Participation, migration, productivity, health and the design of institutions change the consequences of aging. Nor does a fertility rate supply a population forecast on its own. Cohort size, mortality and migration remain part of the demographic arithmetic." ], "groups": [ { "figures": [ 23, 24 ], "layout": "sequence", "paragraphs": [ "The approximate replacement benchmark in the fertility chart is a demographic reference, not a policy target or a guarantee of population stability. It concerns a long-run reproduction concept under assumptions, whereas the plotted total fertility rate is a period measure. The distinction prevents a useful reference line from becoming an explanation of the next year's population change.", "Environmental comparisons require an equally explicit perimeter. Territorial emissions are attributed to where production occurs. Consumption-based measures also account for emissions embodied in trade. A reduction in domestic emissions cannot, on its own, establish an equal reduction in the footprint of everything residents consume." ] }, { "figures": [ 25, 26 ], "layout": "sequence", "paragraphs": [ "The carbon series measures territorial carbon dioxide per person, excluding land-use change. It is not a measure of every greenhouse gas, a consumption footprint or a comprehensive environmental score. Per-person figures make differently sized economies more comparable, but they still leave production structure, energy systems and climate outside the explanation.", "The connected output-emissions chart preserves both observed coordinates. Movement toward more output and fewer domestic emissions is visible where it occurs. It does not identify how much resulted from technology, energy substitution, structural change or activity moving abroad. Calling the path a fitted decoupling frontier would add a model that has not been estimated.", "These qualifications do not make the descriptive progress meaningless. They identify the boundary within which it is true, and the additional evidence required before extending the conclusion to consumption, global emissions or a particular policy." ] } ], "takeaway": "A constraint changes the question policy must answer. It does not supply a forecast or assign responsibility by itself." }, { "id": "dependence", "number": "06", "label": "Bound together", "title": "A relationship can be large for both sides, and indispensable to only one.", "deck": "Return to Canada. This time the denominator is the export book.", "paragraphs": [ "In the consistent 2025 goods-trade snapshot, 72.5% of Canada's merchandise exports went to the United States. Canada accounted for 15.4% of US merchandise exports. The same commercial relationship therefore occupies very different positions in the two economies.", "The directional chord separates exports in one direction from exports in the other. Ribbon color identifies the exporter, the arrowhead the destination, and width the value in current euro billions. Region arcs reflect the displayed incoming and outgoing links, not shares of world trade. Domestic trade, services and investment are outside this view." ], "groups": [ { "figures": [ 27, 28 ], "layout": "sequence", "paragraphs": [ "Volumes and shares do different work. The flow widths retain one common monetary scale, so the small Canadian export system is not silently enlarged to match the others. The printed shares then identify how concentrated each export book is. A large volume describes commercial scale; a large share describes exposure to a destination.", "The EU denominator is extra-EU merchandise exports. Trade among members is excluded because the question concerns the bloc's external relationships. US-bound goods represented 21.0% of that extra-EU total. Canada's EU-bound goods represented 5.5% of Canadian exports. Neither percentage includes services or investment.", "Concentration is not a direct estimate of the cost of a trade restriction. Substitution possibilities, sector composition, contracts, logistics and the value added retained at home all matter. The aggregate export book identifies an exposure; it does not tell us how quickly a factory can replace a customer." ] }, { "figures": [ 29 ], "layout": "feature", "paragraphs": [ "Even the value of a single trade direction depends on a reporting perspective. Exporter and importer records can disagree because of valuation, transit, re-exports, partner attribution, timing and revision. The main view keeps exporter reports. The mirror comparison shows the alternative records rather than averaging them into a number that no statistical agency reported.", "The common European Commission factsheet snapshot is dated May 20, 2026 and concerns goods trade in 2025. Later revisions may differ. A mirror gap is not a trade balance, an uncertainty interval or evidence of misconduct; it is a reason to keep the reporter and the statistical definition visible." ] } ], "takeaway": "Changing a holiday destination and changing a production system are different kinds of adjustment. The evidence should say which one it describes." } ], "conclusion": [ "The comparison does not end with a winner because the measures did not ask one common question. America's resource advantage invites a harder examination of social outcomes. Europe's aggregate invites scrutiny of internal differences and the capacity to sustain future investment. Canada's position invites scrutiny of per-person growth and the concentration of commercial relationships.", "For policy, the useful next step is to name the outcome, choose the denominator and identify the evidence that could distinguish competing explanations. A spending ratio can start that inquiry. A historical path can constrain it. A country ranking cannot complete it.", "Carry a comparison away, but keep its boundary attached. The strongest version of the good life is not the one that wins the most columns. It is the one whose claims survive the change from a national story to the people, institutions and dependencies that story is meant to describe." ], "boundaries": [ { "title": "Observed and derived", "text": "Frozen official historical series and deterministic transformations. IMF historical entries can include staff estimates; mortality indicators include official modeled estimates. No forecasts or policy scenarios are presented as outcomes." }, { "title": "Different common years", "text": "Most annual series end in 2024. The joint health comparison and R&D end in 2023; public education spending uses 2021; PISA and merchandise trade use 2025. There is no all-current synthetic year." }, { "title": "An indicator-specific Europe", "text": "Fixed EU27 membership. Published WDI aggregates where defensible; GDP-weighted national-government expenditure; a 26-member median for education spending; individual countries, never a fabricated student aggregate, for PISA." }, { "title": "No causal tournament", "text": "Country differences do not identify the effects of spending, immigration, regulation or institutional models. Alternative explanations and unobserved composition remain beside the affected claims." }, { "title": "Different distributions", "text": "Country dispersion is not household dispersion. PISA percentile ranges describe students; confidence intervals describe uncertainty around means. The two are not interchangeable." }, { "title": "A partial account of the good life", "text": "Median consumption, wealth, leisure, housing affordability, access, subjective well-being, services trade and causal institutional effects are outside the harmonized evidence here. They are unavailable for this comparison, not zero." } ] }