The article compares GDP per person in Canada and the US by looking at differences in productivity and work intensity. It shows that using different data sources for hours worked can lead to big errors in understanding these differences. In 2000, comparing hours worked from employer surveys to household surveys made it seem like 72% of the GDP gap was due to productivity, when it was actually only 36%. This highlights the importance of using accurate data when comparing economic performance between countries.