Three versions
of the good life.

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.

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.

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.

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.

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.

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.