In 1998, the STScI was assigned to be the Science & Operations Center for the Next Generation Space Telescope (NGST), which has now been renamed the James Webb Space Telescope (JWST.) At the time, JWST was scheduled for a 2007 launch, meaning that the Institute would have to start development for ground operations within a few years.

The STScI organizational structure was designed to support a single large project, the Hubble Space Telescope. It was already clear in 1998 that the structure did not lend itself easily to supporting two large projects, one in operation and one in development, without some modifications. As Director, Steve led the senior management team that would create a new organizational structure to handle Hubble, JWST, and allow sufficient flexibility to manage the operation of smaller projects that made up an increasing part of the STScI work.

The Institute's Management Council (MC), consisting of all senior managers including the Director's Office, had held a series of offsite retreats to design a new organizational structure that would address the changing environment. The retreats took place in May through August 1999 involving the help of a management consultant in organizational structure. The MC divided into three groups and examined three different approaches to organizational structure based loosely on Activities (i.e. structure based on work activities common to both projects), Projects (a traditional 'stovepipe' structure in which the HST and JWST staff were separate organizational elements), and a Matrix (in which teams were created from Divisions of skilled labor to carry out work specific to mission needs.) The MC did not reach consensus on which of these approaches was optimal for STScI.

Steve ultimately selected the Matrix approach for its flexibility in response to changing mission needs, common to large space projects, and then made some modifications to the proposed structure based on recommendations of the Director's Leadership Forum (DLF, see separate web page.) He then carried out the re-organization in two stages, the first put in place in January 2000 and the second put in place in October 2002. Steve held a number of All Hands meetings as well as meetings with individual divisions in the Institute to explain the re-organization strategy. Two presentations that give examples of these meetings are All Hands 99Nov98.ppt and All Hands 02Sep30.ppt.

The first stage created a JWST project office to define the development work at STScI with a small team to propose the work to NASA. The staff resided primarily in one of six large divisions: Hubble, Engineering, Archive & Data, Public Outreach, Computing Services, and Administration. Several other divisions, such as NGST (JWST), Science Policies, and Science, were smaller but carried out important, high-profile work. This first phase of the re-organization recognized that there was still very little work for NGST, and so it left most of the work teams supporting Hubble in place with minor changes in the reporting structure, mostly at the senior management level. It also created the position of Chief Information Officer in the Director's Office to oversee the Institute's decisions on choices of software, hardware, and computing development that would have long term impacts on the projects and STScI infrastructure.

The second stage moved STScI to a true matrix structure where there were two large projects (HST and JWST) and several smaller projects (MAST, FUSE, Kepler, etc.) that needed to be supported with the full range of skills of the staff: science, engineering, computing support, business services, and outreach. This structure, still in place today, has the staff residing primarily in five large divisions: Engineering and Software Services (ESS), observatory instruments (INS), Operations and Data Management (ODM), the Center for Processes and Technology (CPT), and the Business Resource Center (BRC, formerly Administration.) The Science and Science Policies Divisions were left alone as was the Office of Public Outreach. These divisions manage the teams that work for the projects.

The project offices for HST and JWST consist of a few people who oversee the work of the teams and make sure the work priorities accord with the project goals. For very small projects, one individual may have the responsibility to make sure the work is organized properly by the divisions.

The organization has run well since October 2002. It continues to receive high ratings for its work from NASA, and it has reacted well under the stress of several crises: the loss of the STIS instrument, the delay of JWST, the cancellation of the final servicing mission and its subsequent re-instatement by NASA.