The decomposition of changes in a country’s carbon emissions into contributions from changes in the sector-level carbon intensity, value-added share of sectors and level of GDP illustrates their relative importance for observed emissions. Using data from 34 sectors in 39 countries covering 1995-2009, I find that shifting composition of output towards low-carbon sectors and improvements in sector-level carbon intensity were instrumental in constraining emissions in most countries. In contrast, increases in economic activity pushed emissions up in all countries. These observations suggest structural change can be as important as cleaning up carbon intensive sectors and motivate a closer look at high (HCI) and low (LCI) carbon intensity sectors. I document the large cross-country variation in the average carbon intensity of HCI and LCI sets and of individual HCI and LCI sectors. HCI sectors tend to (i) account for a smaller share of employment; (ii) be more capital intensive; and (iii) employ a workforce with a lower average skill level. In the full sample, and in subsamples by level of development, employment declined in HCI sectors and increased in LCI sectors with its composition shifting towards high-skilled workers in both. Capital intensity growth was faster but multifactor productivity growth was slower in HCI sectors. Moreover, the pace of change was typically greater in the LCI sectors of the developing country subsample.

Keep in touch with the Grantham Research Institute at LSE
Sign up to our newsletters and get the latest analysis, research, commentary and details of upcoming events.