On October 13, 2010, President Barack Obama signed the Plain Writing Act of 2010 into law. It requires federal government agencies to use plain writing in all "covered" documents "the agency issues or substantially revises" (Sec. 4 (b)). The goal of the Plain Writing Act is to "enhance citizen access to Government information" (Introduction) and improve government operations and accountability "by promoting clear Government communication the public can understand and use (Sec. 2). This dissertation examines what plain language, as the Act defines it and the U.S. federal government (USG) is implementing it, means and does at the various levels of language policy -- institutional, writer, and user. I argue that "real" plain language policy differs from the policy documented by the Plain Writing Act of 2010. While plain bureaucratic writing can help to make government documents more understandable for users, plain writing alone cannot achieve the Act's goals. Plain writing is a style of writing. As such, it is not only contingent, but it is also subjective and based on preference, which is impossible to legislate. Hence, plain bureaucratic writing is and does different things at different levels of language policy. Moreover, institutional- and writer-level representations of plain bureaucratic writing are at odds with user representations in many respects. Plain bureaucratic writing for USG agencies and federal writers is another way to describe good writing in the tradition of Edited American English and "fixes" the problem of bad government writing. At the user level, understandable writing is plain writing, regardless of whether it adheres to the principles of standard written English or the plain style. Plain language legislation does not affect the work of most writers in the study or their ability to do it. Nonetheless, user participants generally prefer plain language, reporting that they are more inclined to do what a government document intends for them to do when they understand it. Efforts to enhance government communication should focus on usability instead of plain language since usability is a better measure of the extent to which plain bureaucratic writing impacts the textual government-citizen interaction. / Ph. D.