Weirdly, when I pop the data into a linear regression program, it pretty clearly shows an upward trend. Additionally, the article preferred (strongly) to point to non-peer reviewed articles and not actual data.