Analyses. Data were recorded and analyzed in Microsoft Excel and StarOffice Calc. Service provider data were remitted in varied forms, including individual costs per day, total cost per service, cost per person, or total budget. Costs were divided to obtain a per person, per service figure. These figures where then multiplied by reported service use by people experiencing homelessness and then multiplied by the percentage of people experiencing homelessness that use the particular service. If a per person, per service figure was unattainable, the total budget or total service cost was divided by the homeless count to determine a best estimate figure. When data were insufficient for an estimate or specific data were unavailable, figures where sparingly used from an existing study with the most similar region, population, and goals avaiable – Clasen's (2006) study of the hidden costs of homelessness in Durham, NC. Each figure will be discussed in the results section.