Housing patterns have changed over the past 5 years as the housing boom has spread outward from many metropolitan areas. In the US, unlike the UK and other European countries, the workforce is much more adaptable to changes in employment opportunities.
On a macro level, as employment opportunities show greater strength in one region over another, we usually see a gradual shift in population to those areas. On a micro or local level, commuters have been willing to lengthen their travel time in search of housing that matches their income levels. This is encouraged by the expectation of future price growth, one of the key consumer components of the current boom.
Two patterns have emerged:
Residents have remained in the city as urban areas have seen a renaissance as represented by the tight housing supply and increasing demand.
Home buyers have been willing to extend their commutes as housing prices rise. 
I did my own informal study in 1990 when my family and I were looking for housing in the Westchester and Fairfield Counties (long considered commuter suburbs for New York City). I found that the average price of housing at that time decreased about $100,000 for each 10 minutes past the first express stop as you move away from the city.
Academia has long attempted to quantify the relationship between commute time, wages and housing costs.
Here are a few examples:
University of Chicago [Note: PDF]  looked at the response of housing rental rates, wages, employment and population density can be evaluated in response to changes to highway capacity.
Iowa State did a study on The Effect of Housing Prices, Wages, and Commuting Time on Joint Residential and Job Location Choices [Note:PDF]  analyzed how these factors determined where people lived.