Has average job tenure declined over time?:
The More Things Change, The More They Stay the Same: Trends in Long-term Employment in the United States, 1969-2002, by Ann Huff Stevens, NBER WP 11878, December 2005: Abstract This study considers whether there has been a decline in the attachment of workers and firms in the United States over the past several decades. Specifically, it compares snapshots of job tenure taken at the end of workers' careers from 1969 to 2002, using data from the Retirement History Survey, the National Longitudinal Survey of Older Men, and the Health and Retirement Study. The primary finding is one of stability in the prevalence of long-term employment relationships for men in the United States. In 1969, average tenure in the longest job for males aged 58-62 was 21.9 years. In 2002, the comparable figure was 21.4 years. Just over half of men ending their careers in 1969 had been with a single employer for at least 20 years; the same is true in 2002. This finding is robust to adjustments for minor differences in question details across data sources and for educational and retirement age changes over this time period.
From the beginning of the paper:
A great deal of attention has been paid in recent years to the issues of job stability and job security in the United States. Many studies have attempted to determine whether there has been deterioration in the prevalence of long-term, stable employment relationships in the United States during the period from roughly the mid-1970s to the present. Most of these studies have found either (1) no robust evidence of significant changes in various measures of job tenure, job stability or job security or (2) indications of relatively small increases in job turnover, particularly during the early 1990s. In contrast to the findings of these studies, there remains a powerful conventional wisdom that the U.S. has experienced widespread, substantial declines in expected job security or stability. ... Further, when asked directly, workers themselves appear to be more worried than in previous years about the risk of separating from their employers. (Schmidt, 2000). There is, however, a striking lack of solid empirical evidence to support these claims. Even in cases where careful studies have shown some decline in job tenure or increased turnover, the question remains whether the magnitude of documented changes can justify claims of major shifts in employment relationships often found in the popular press.
There is other research, e.g. see here, showing that measures of insecurity such as the variance of income have risen in recent years. Thus, even if these results hold up to further scrutiny, they do not prove that worker's perception of increasing economic insecurity is illusory.