Europe vs the US: reconciling Krugman, Garicano and Ackerman
Javier López Prol
Department of Economics, Yonsei University - Mirae, South Korea
14-05-2026
The controversy
There is an intense debate about the relative economic performance of the United States and Europe. Garicano and coauthors argue that European living standards and productivity are stagnating compared to the US (P. Garicano and Garicano 2026; L. Garicano and Kooi 2026), while Krugman claims that both are evolving at the same pace when accounting for prices and hours worked (Krugman 2026a, 2026b, 2026c), and Ackerman even suggests that “Europe’s productivity is outpacing the US” when using time-consistent price deflators (Ackerman 2026). While their narratives comprise many elements, I will focus here on the minimum that is economically tractable: living standards and productivity.
These authors use different concepts, units of analysis, periods and indicators, so it is difficult to directly compare their arguments and numbers. I’ve tried to reduce it to the lowest common denominator to reconcile positions and derive some conclusions. First, I will define living standards as GDP per capita in purchasing power parity (PPP), and productivity as GDP per hour worked in PPP. I will take the regional definitions proposed by Garicano (L. Garicano and Kooi 2026), since Western and Southern Europe (WSE) show a significantly different behavior from Eastern Europe (EE)1. I will take two data sources: World Bank (WB) data, commonly used by Krugman and Garicano, and the Penn World Table (PWT) as suggested by Ackerman, for the period 2000-2023 (latest available in the PWT).
What’s the picture today?
Figure 1 shows the static comparison of WSE and EE to the US in 2023. The gap between the US and Europe is large when converting currencies at market exchange rates, but declines when adjusting for differences in price levels across countries (PPP). When further adjusting for hours worked, the gap between the US and WSE declines to between 6-11% depending on whether we use WB or PWT data, and between 42-55% for EE. The sources differ on the magnitude of the gap between the US and Europe, but they broadly agree on the direction: the US is richer and more productive than WSE, and much more than EE. The main disagreement is about the evolution across countries and over time, and for that we have to discuss the indicators.
Figure 1. Europe vs the US in 2023
Convergence or divergence?
International Economics 101 teaches that we have to adjust price levels across countries with PPPs when we are doing international comparisons, and we have to adjust for inflation by keeping prices constant at a base year when comparing real growth over time. The problem arises when we want to adjust for both things at the same time.
Garicano’s approach (P. Garicano and Garicano 2026; L. Garicano and Kooi 2026) is to use constant PPPs. This entails adjusting for differences in price levels across countries at the base year, and then extrapolating national growth rates using national deflators. While this preserves valid real growth internally (within country), it presents two problems. First, keeping PPPs constant at the base year ignores changes in relative prices differentials across countries over time. Additionally, national deflators are inconsistent across countries, as each national statistical office has their own criteria, as Ackerman documents (Ackerman 2026). This creates inconsistencies in comparisons of countries over time. The opposite approach, suggested by Krugman (Krugman 2026a, 2026b), is to use current PPPs. Then prices across countries are updated every year, but the resulting series is not deflated for inflation, so it is not a clean measure of real growth.
This problem is not new (Deaton and Heston 2010; Aghion et al. 2019, 2018) and there have been efforts to create time-consistent indicators across countries. The best available effort in this direction, to my knowledge, is the Penn World Table (Groningen Growth and Development Centre 2025), which offers two main innovations. First, PWT separates expenditure-side (e) and output-side (o) measures of GDP. The expenditure-side series is the natural benchmark for comparing living standards, because it is closer to what residents can purchase. The output-side series is the natural benchmark for comparing productive capacity, because it is closer to what the economy produces at international prices. PWT also provides chained PPP series, which are better suited than either current PPPs or fixed-base constant PPPs alone for comparing countries over time. (Feenstra, Inklaar, and Timmer 2015) explain why PWT separates expenditure-side living standards from output-side productive capacity and why chained series are preferred for cross-country comparisons over time. More importantly, the PWT produces a new series of real GDP with “chained”, rather than purely constant or current PPPs. The chained PPP series link PPPs across revisions to improve real growth consistency over time, while also adjusting for differences in price levels across countries.
Since living standards are a measure of demand-side purchasing power, and productivity is a measure of supply-side productive efficiency; and given that chained PPPs are the best effort to adjust for both price levels across countries and inflation over time, I will use PWT GDP per capita from the expenditure approach and chained PPPs to measure living standards, and GDP per hour worked from the output approach and chained PPPs to measure productivity. Then I will compare these series with the constant PPPs proposed by Garicano and the current PPPs used by Krugman.
Figure 2 shows the change in living standards (GDP pc) and productivity (GDP per hour worked) for WSE and EE relative to the US between 2000 and 2023. The PWT chained PPP series (solid blue) are used as the preferred benchmark, and are compared for each concept and region with WB constant (dotted orange) and WB current PPP (dashed green) GDP indicators.
One conclusion is clear: EE has strongly converged to the US in both living standards and productivity. The picture is more ambiguous for WSE, hence the importance of the right metric choice. Living standards evolved in the same way in WSE and the US, according to the PWT GDP(e) at chained PPPs. However, WSE converged to US living standards at current PPPs (Krugman’s perspective), and diverged at constant PPPs (Garicano’s perspective). Similarly, productivity diverged in WSE compared to the US according to both the PWT GDP(o)/h at chained PPPs and also at constant PPPs (Garicano), while productivity maintained its relative level to the US at current PPPs (Krugman). Here is the source of the disagreement.
Figure 2. Change in GDP per capita and per hour worked in WSE and EE between 2000-2023. WB constant PPP is rebased to 2023.
Where is the bias coming from?
Using PWT chained PPP series as the preferred benchmark for time-consistent cross-country comparisons, we can decompose the bias of the WB current and constant PPP series relative to the PWT chained PPP series. This is important because the WDI series are more commonly used in public debate and policy discussions, but they can give a very different picture from the PWT series. There are four potential sources of bias:
WB - PWT current PPP source gap captures the difference between WB current PPP GDP and the comparable PWT current PPP GDP concept (green).
WB constant-current wedge captures the additional difference created by using WB constant PPP instead of WB current PPP (red).
PWT current-price revaluation captures the difference between repeated current-PPP PWT levels and the PWT chained real series (blue).
PWT output-expenditure bridge captures the difference between PWT output-side and expenditure-side GDP. This term matters for productivity, but not for GDP per capita (purple).
Figure 3. Bias of WB current and constant PPP relative to PWT chained PPP. Bars decompose the bias; points show the net bias.
Figure 3 shows that the WB current PPP tends to favor European convergence relative to the PWT chained PPP benchmark, whereas the WB constant PPPs indicate more European decline compared to the US. The current PPP bias is explained by the WB-PWT current PPP gap and the PWT revaluation, while the constant PPP bias is mostly explained by the WB constant-current wedge. The output-expenditure bridge only matters for productivity because both GDP pc indicators use the expenditure approach.
Verdict
This harmonization suggests that much of the disagreement between Krugman, Garicano, and Ackerman comes from the choice of PPP concept and from whether the object of interest is living standards or productivity. For WSE, PWT expenditure-side chained PPPs imply that GDP per capita has remained broadly stable relative to the US, which is closest to Krugman’s living-standards interpretation. PWT output-side chained PPPs imply that GDP per hour has declined relative to the US, which is closer to Garicano’s productivity concern. Ackerman’s broader point is that chained PPP series are better suited than either current or constant PPPs alone for comparing countries over time.
References
WSE includes Austria, Belgium, Switzerland, Germany, Denmark, Spain, Finland, France, the United Kingdom, Greece, Iceland, Italy, the Netherlands, Norway, Portugal, and Sweden. Ireland and Luxembourg are excluded because multinational profit shifting and small financial-center effects distort GDP. EE comprises Bulgaria, Czechia, Estonia, Croatia, Hungary, Lithuania, Latvia, Poland, Romania, Slovakia, and Slovenia. All aggregates are constructed by summing GDP, population, and hours first and then dividing.↩︎