The statistical program to create estimates for the paper entitled “Cost Shifting To Jails After A Change To Managed Mental Health Care” (Domino et al. 2004) contained a typo in the inflation adjustment for year 1994 data, making the pre-managed care spending by King County Mental Health too large. This result affected the estimates presented for King County Mental Health, but does not affect the policy conclusions. The following table contrasts the published estimates with the corrected estimates:

Section in HSR PaperWhat Remained the SameWhat Changed with Revised Estimate
Table 1: Summary Statistics for dependent variables% with any use, % with monthly useMonthly expenditures for users drops from 381 (SD 649) to 330 (SD 463)
Table 2: Summary Statistics for independent variablesEverythingNothing
Table 3: LPM on any use of servicesEverything including KCMH model, because change does not affect the number of usersNothing
Table 4: Level of use for usersSignificance of all results in KCMH modelLevel of effect (MC × Medicaid interaction) increased from −203 to −124); other variables, especially MC also change in magnitude (−316 to −239)
Simulation of annual spending estimate (not in table, but reported on pp. 1395–6)All other estimates, and the fact that spending dropped. Policy conclusions remained the sameThe level of the drop decreased from $4.6M (6.3% of budget) to $2.2M (3% of budget)


  1. Top of page