Over the last two years, the American Community Survey has made changes to its data collection operations. On October 1, 2012, the ACS reduced the scope of the Failed Edit Follow-up operation, resulting in substantially fewer incomplete self-response returns going to follow-up. In addition, we added a new mode of data collection starting with the January 2013 ACS panel, allowing respondents to complete the survey over the Internet. Because of these changes, we expected to see differences in item nonresponse. We compared item allocation rates prior to the changes to item allocation rates after the changes were made and found the reduction in follow-up resulted in higher allocation rates. The addition of the Internet mode helped lower the allocation rates for many of the items in the housing section of the survey, however the new mode resulted in even higher allocation rates for the items in the detailed population section of the survey.