Thanks, Jim, for bringing 'development' back to the a discussion about the 'development' of civic identity. A somewhat related point is that an identity search, regardless of identity type, does not start until a young person reaches adolescence but is not necessarily resolved by any specific age (not to step on Erikson's toes - at the very least, the social construction of adolescence, in this day and age, extends past 18 or 19 for many youth.). In other areas, such as ethnic identity, research suggests a lifespan development model in which the identity search is a continuous process, influenced by life experiences.
In an interesting conversation with Lew Friedland yesterday we discussed how the context of an 18 or 22 year-old during Erikson's time was much different than the experiences now. So, for instance, an 18 year-old high school graduate would in a sense need to resolve their identity search, choose a job that they would hold for decades, move into a community where they would live for those decades, marry and have children relatively early. If a person didn't resolve their search, then they'd be considered abnormal by societal standards and would have few opportunities to pursue alternative paths in their identity search (hippies, for instance, were considered by Erikson to be pursuing an extended adolescence). Today, though, regardless of education, young people are encouraged to try on many hats throughout their 20's (and even into their 30s?).
When we think about civic identity, consistent with Jim's thoughts, we shouldn't necessarily be concerned with the influence solely of schools (or any other single institution) at one given time point.