For the first time since April 2007, our statistical indicators show that a recovery is more likely than a contraction.

After increasing for four months in a row, the percentage of primary leading indicators expanding stands at 78 (seven out of nine indicators with a discernible trend). The cyclical score of leaders, which is based on separate purely mathematician analysis, has increased to 54, and we revised the last month’s score from 46 to 51. When the percentage of leaders expanding and the cyclical score both rise above 50, expansion is more probable than contraction.

Despite the relatively encouraging picture painted by the leaders, the employment situation remains bleak. It suggests that the upcoming recovery might turn out of be jobless, similar to the recoveries in 1992 and 2001.

Initial claims for state unemployment insurance, for example, decreased in September to approximately 540,000 claims per week, significantly below the peak of 658,000 reached last March. The series, inverted for analysis, is now appraised as clearly expanding.

The decline merely means that the labor market is no longer getting worse quite as rapidly. But it is not getting better: The unemployment rate has increased from 9.7 percent in August to 9.8 percent in September. In order for the unemployment rate to decrease, initial claims have to fall considerably—to below 350,000 per week.

One reason this recovery may be jobless is because of reallocation of employment among different sectors of the economy. This usually results in longer unemployment than would temporary layoffs within a sector.

For more than 50 years, employment in goods-producing industries (mining, construction, and manufacturing) has been shrinking as a percentage of total non-farm employment. In the 1950s, almost 40 percent of all workers worked in these industries. Currently less than 15 percent of workers do. Technological changes that make manufacturing processes more efficient are responsible for this trend. These changes cannot be stopped.

Since employment in the goods-producing industries is trending down over the long term, there will not be enough jobs to employ all workers currently laid off in those industries. Some of them will have to find work elsewhere. But changing an occupation usually requires more time and training than would a simple return to an old job or industry. When this kind of job reallocation occurs on a large scale, it takes longer for unemployment to go down after the expansion begins.

Read a general description of our indicators and methods of analysis.

• Receive a detailed monthly analysis of business conditions, plus articles and books on a wide range of economic issues. Learn more about the benefits of Membership.