The NY Times has an article that provides some details about the decline of carpooling since the 1970s. The above graphic is from that article. Overall the mode share of carpooling has declined from 20% in 1980 to 10% in 2009.* Certainly the changing metropolitan form has much to do with the relative decline as jobs are now spread throughout the region, and I'm glad they mentioned the differences in carpooling by immigrant status. But some other factors not mentioned deserve notice, including the growth of women in the workforce. Couples used to be the largest share of carpoolers, but now men and women within one household are less likely to work in the same locations making carpooling difficult.
Other factors at play are as the networks of carpool lanes have grown (though there are few in New York City) regulations restricting "slugging" or informal carpools often through tolled bottlenecks have increased. Perhaps slugging should be allowed in more areas.
It is hard to view the massive investment in carpool lanes as much of a success in light of the dramatic decline in carpooling. The positive outcome is the United States has a robust network of lanes that can easily be converted from HOV lanes to High-Occupancy Toll (HOT) lanes. These conversions are happening in many metropolitan areas, and are likely an overall improvement for congestion though they are not universally loved.
The expansion of carpool lanes has not fulfilled expected outcomes for commuting behavior. Perhaps we can alter the use of these lanes to improve transit, congestion and travel choices. The lanes certainly go to show that we "waste" money and make bad choices with transportation investment across modes and that cars are no exception, yet I don't see the same level of scrutiny toward $ billions spent on highway expansion through carpool lanes that I see toward rail transit--and I include myself in this critique.
One last point is that towards the end of the article the authors note that transit use is growing while carpools are declining:
As car-pooling has continued to decline, mass transit use has increased in the past decade. In the Washington area, it represents about 14 percent of commuters, compared with 11 percent in 2000, according to the data.
While true, this is misleading as the initial bases from which these shifts are measured are so different. Transit is was and is low. It may be growing but it is still only half of the mode share of carpooling (5% v 10%). Maybe none of this matters much as commuting is only about 20% of travel, but transit is not always well designed for non-work trips while carpools are much more popular for non-commute trips as people go out and about with friends and family.
*Technically this is an average of 2005-2009 rather than a snapshot as were previous data years.