Due to the poor performance of US students in international math and science tests, many authors worry that the US lead in science is in jeopardy. A recent study by Chen and Luoh ("Soc Indic Res" 96: 133-143, 2010) challenged this pessimistic view by delinking test performance and labor force quality. It was found that measures such as researchers per capita or scientific journal articles per capita can better account for economic performance. Because it takes time for education and R&D to yield returns, this project replicated the preceding study using a time-lag structure. Although the results partly concur with previous findings, it is important to point out that GDP might depend on multiple variables acting together. Instead of solely focusing on reforming STEM education as partly measured by international test scores, improving the R&D infrastructure and openness in trade could also be crucial to the future of the economy.