BRYAN CAPLAN
May 7, 2013
Keynesian Bets: What's Out There
May 6, 2013
Keynesian Bets Bleg
May 6, 2013
The Pyramid of Macroeconomic Insight and Virtue
May 2, 2013
A Natalist Provision
May 1, 2013
I Was a Teenage Misanthrope
DAVID HENDERSON
May 5, 2013
John Thacker on Vaccinations and the Sequester
May 3, 2013
Chef Rudy's Virtues Project
May 2, 2013
My take on Reinhart and Rogoff
May 1, 2013
Medicare Kills a Program


Both ACS and March CPS include very detailed variables on education, income, and health insurance status, including whether the health insurance is provided by the employer or through some other source. The microdata is accessible through ipums.org.
This isn't quite what you're looking for, but BLS can tell you health insurance incidence by occupation (see the Employment Cost Index section), and many occupations have minimum education requirements, so failing a better answer you could get there indirectly.
You should use the SIPP; it has this information. For example, table 2 on this page (http://www.census.gov/sipp/sb94_06.pdf) resembles what you are looking for (health insurance lapses by education level). Conveniently, there is a simple R interface to SIPP data with an tutorial at http://www.asdfree.com/2013/02/analyze-survey-of-income-and-program.html
If you and/or your graduate assistants are unfamiliar with R or the SIPP, let me know and I will write the R program to compile the table for you and send you the code.
A comparison of the three available census sources is shown here. http://www.census.gov/hhes/www/hlthins/about/index.html