An Input Adjustment Method for Challenging Privatization: A Case from Michigan Prison Health Services

Comparing Relative Value

There is near universal agreement throughout the voluminous literature on privatization that successful private contracting depends on whether agent effort can be efficiently channeled toward a publicly defined mission. In general, proponents of privatization place trust in the ability of public agencies to design effective tendering processes that leverage the salutary effect of competition and the instrumental power of service contracts to secure agent compliance, and in the capacity of public agencies to spot moral hazards and take corrective actions (Savas 2000). Skeptics question the practicality of this delivery model given the innate obstacles to implementing managed competition: substantial transaction costs, monitoring challenges, transition costs and bidder scarcity that impede the replacement of subpar agents, or political influence that distorts the objectivity and motives of decision makers. ² Closely examine privatization cases, claim skeptics, and reported savings are often an illusion attributable to change in output or are the product of questionable accounting practices (Sclar 2000). Arriving at a weighted consideration of price and substance (i.e., relative value) is central to these debates, and, from an administrative perspective, is essential for assessing the extent to which a private agent fulfills a public mission.

One risk to privatization is that the contractor dilutes the scope or quality of the original public function. Pressure comes from two sources: (1) the effort to lower costs for a competitive bid and (2) the drive for profits. The first happens during the bidding phase when prospective vendors vie to secure an initial contract. “Low balling” is the strategy to submit an unrealistically low bid price to win a contract, with the expectation that additional revenues can be coaxed from political channels afterward. The second form of pressure relates to the postcontract operational adjustments by a vendor to balance contract compliance with revenue. Once ensconced in a system, vendors begin receiving information on costs and, where feasible, incrementally trim higher cost areas to gain profit. For most public services, this involves discouraging labor-intensive activities, which in turn can alter the character of a service.

Such behavior is hard to control with even the most diligent agency monitoring. Administrators making decisions on the public’s behalf should welcome a method for forecasting the change in substance prior to a make-or-buy decision. For outsiders who wish to challenge privatization, these tendencies are also problematic because it is usually quite easy for a private bidder to lower the cost of a public function by providing less substance. What the outsider must do is expose the false comparison and adjust costs commensurate with substance to arrive at relative value. There are two approaches for locating unequal substance: compare expected outputs (i.e., the difference between private and public in terms of goods and services rendered) or compare inputs (i.e., the difference between private and public in terms of labor, capital, and supplies). This study presents an example of the second approach.

Some Characteristics of Prison Health Care

Distinguishing prison health care from clinical practice in free society helps to convey the specialized role of a prison health provider. Prisons are communities of integrated functions that share a common purpose of inmate custody and security. MDOC’s formal mission to “create a safer Michigan through effective offender management and supervision in our facilities and communities while holding offenders accountable and promoting their rehabilitation” draws all job titles into custody and security roles. Thus, beyond occupational skills, workers must know prison rules and procedures for security and safety, including what actions to take during prison conflicts. Health professionals must guard their interactions with inmates; rules exist for ministering treatment, touching, talking, tool inventory, and so forth, which have no parallel in free practice. Acquiring this knowledge takes months to learn and years to become habitual.

An equally relevant environmental factor is a daily schedule featuring long, dull stretches of time interrupted by supervised routines. Inmates try to break the boredom, for instance, by complaining of phantom maladies (e.g., chest pains) in order to persuade the acting physician to prescribe a health screen at a regional hospital. Manipulation of mental health care also occurs as prisoners seek to lift cell restrictions or earn positive parole recommendations. Much more so than in free society, prison professionals are tasked with distinguishing between false and genuine health claims.

Client prisoners have intense health care needs. The average prisoner lacked access to adequate health care prior to incarceration, and from poverty, drugs, and risky behavior, inmates often arrive with serious physical health problems. Once inside, close captivity raises the odds of viral and bacterial epidemics (Jacobi 2005). Moreover, mental health needs within the prison are significantly more intense than in the free population (Watson, Stimpson, and Hostick 2004). As of 2012, approximately 20 percent of Michigan’s inmates required mental health treatment as a condition for rehabilitation. Furthermore, long-term sentencing has resulted in an aging inmate population, increasing the demand for geriatric and hospice services. Prisons must therefore provide the complete gamut of physical, mental, and dental care for a full range of adult demographics.

Prisons are condensed arenas of litigation. There is no “market” for health care in a prison; the incarcerated have little choice but to accept what the state provides, which is why it is unconstitutional to deny inmates care (McDonald 1999). Lawsuits for medical negligence can reach millions of dollars. Even if the state escapes liability, untreated physical and mental health conditions become a community problem when the inmate reenters society (Jacobi 2005). For public health and legal and moral reasons, the state has a vested interest in ensuring effective inmate care.

Privatization’s Ideological Lock on Michigan

Michigan experimented with privately delivered prison health care in the 1990s through a pilot project at five facilities, chosen because their inmate populations were younger than state average and health problems were less severe. In 1995, Wexford Health Sources won the bid to provide physical health services at the sites. In 1996, the MDOC commissioned an external review of the pilot, concluding privatized health care was higher cost than publicly delivered care (Health Management Associates [HMA] 1997). The findings provoked a rebuttal from the CEO of Wexford, which led to concessions in the analytical test. The most substantive was to subtract from the Wexford contract cost the compensation for a health unit manager, a public MDOC monitoring position based on the claim that an on-site monitor would be unnecessary with full privatization. Even after this adjustment, the public system was more cost effective. In their assessment of quality, the review found that the Wexford sites had worse audit scores than comparable public sites and a high turnover of staff. Health-care-related lawsuits were more frequent under Wexford, but this was attributed to variation in prisoner populations (HMA 1997).

One notable difference between the Wexford and MDOC sites was the utilization of off-site specialty care. In the managed care pilot, Wexford had to incur costs for off-site specialty care, and in obedience to these incentives, far fewer prisoners in Wexford locations were allowed access to health services outside the prison walls: a pattern that repeats in later MDOC privatization experiments, where contractors will deny care to prisoners under a managed care model when they are incentivized to do so.

Wexford exited in 1998, but private contracting continued through a contract awarded to United Correctional Managed Care (UCMC). Privatization in this instance was not for on-site clinical care but for off-site specialty care. Under the terms, UCMC was given a fixed monthly budget to meet the health care needs of the inmate population; thus, UCMC profited if fewer prisoners received specialty care. Specialty care utilization declined, while complaints about unmet health care needs grew (Office of the Legislative Corrections Ombudsman [OLCO] 2002) and in-house clinical care costs escalated (Firestone and LaRoux 2000). In essence, UCMC was shifting the burden of care to on-site medical staff and inmates in the form of higher risk. Yet it was the failure to pay vendors, for example, hospitals, ambulances, and so forth, that precipitated the contract termination. Within a year, the UCMC contract was transferred to Correctional Medical Services (CMS).

Initially, CMS resumed the role of UCMC to organize prisoner access to off-site specialty services. In April 2000, the CMS contract was revised when MDOC medical service providers (MSP), specifically physicians, physician assistants, and nurse practitioners, became contracted CMS staff (Firestone and LaRoux 2000, footnote 3). Physicians in clinical care render the diagnoses that determine whether a prisoner is eligible for specialty care. With this change, the persons responsible for specialty care referrals, and who could advocate for inmates by challenging denials of specialty care by CMS’s management arm, became CMS contractors. An important check against the profit-optimizing strategy to deny inmate care was removed.

Other policy decisions weakened the monitoring of CMS. For instance, in 2003 the Legislative Corrections Ombudsman, which previously had unfettered access to hear inmate complaints about health care issues, was defunded and closed. As monitoring became increasingly relaxed, CMS gained latitude to pressure contracted medical staff to minimize specialty care referrals and deny or delay necessary care.

A 2007 report by the American Friends Service Committee of Michigan raised concerns about potential violations of prisoner constitutional rights due to inadequate health care (Tripp et al. 2007). The exposition highlighted a hidden cost to Michigan’s managed care model: legal expenses brought by inmates and advocates, ³ exemplified by the Hadix v. Caruso case, which led to a court-appointed health care monitor for Jackson area prisons in 2003. To remedy the problem, the American Friends Service Committee recommended “all competent-appearing primary care medical services providers (MSPs) . . . be converted to state civil service positions” (Tripp et al. 2007, 117). To improve accountability, the state was to “merge all prisoner medical and mental health care into one interrelated management structure, in one department” (Tripp et al. 2007, 118). The report warned against the underlying conflict between inmate care and private delivery, offering that “. . . the state should reconsider a return to MDOC-operated health care rather than contracting with an entity that expects to turn a profit in a field where any such profit comes at a human loss” (p. 16).

Using more circuitous language, a January 2008 report funded by the MDOC reinforced this position. According to the National Commission on Correctional Health Care’s (NCCHC 2008, 5) assessment, the MDOC health care system was complicated, subpar in meeting recognized standards (p. 9), duplicative (pp. 34-35, 39), unproductive (pp. 23-24, 39), and where inmates faced extensive waiting periods to receive basic services and were often denied specialty care (pp. 7-9, 33). Described as “one of the most bureaucratic [we have] ever encountered” (p. 22), efficient practices like telemedicine were underutilized (p. 10), medications were wasted or unavailable (pp. 12-14), and there was a lack of coordination within and across service units. Decentralization of responsibility through contracting was the culprit. ⁴

These problems were not due to weak contract language. As the NCCHC report describes, the service contract had language requiring staffing standards and penalties if CMS failed to meet them. The problem was lax enforcement. Service monitoring was nonexistent; the MDOC did not issue a single monitoring report over a ten-year period even though the contract stipulated regular performance audits (NCCHC 2008, 33). No penalties were assessed when CMS failed to meet contractual obligations.

A recurring theme was the loss of administrative control. In several places, the operational problems were because “providers are not [MDOC] employees” (NCCHC 2008, 6, 23). The MDOC was powerless to remedy low productivity because the MSPs were independent contractors under CMS; at the same time, CMS claimed that they could do nothing to remedy the matter (p. 24). Telemedicine, a cost savings tool that eliminates inmate transportation for certain types of diagnostic care, was underutilized because CMS was required to pay for the technology under the managed care model (NCCHC 2008, 10). Through privatization, the MDOC abdicated control to mandate cost-effective forms of care.

Near the end of the report, the NCCHC urges the MDOC to reconsider adopting a public model of care delivery. Recommendation number 23 was that MDOC seriously weigh the advantages and disadvantages of privatization. But the follow-up strategic plan never considered “the option of the MDOC returning to Civil Service providers” (NCCHC 2008, 51). Instead, the MDOC limited its scope for service restructuring to altering the terms and conditions of a privatized system of care. This most recent foray is consistent with Michigan’s ideological commitment to privatization.

Private versus Public Prison Health Care

The 2012 RFP required all bidders to standardize their pricing based on “per prisoner per month” (PPPM) amounts. Two firms responded to the RFP: Corizon and Centurion. Corizon, which was formed through a merger between CMS and Prison Health Services (PHS), is the largest firm in this industry and had contracts with the MDOC since 2009 to manage specialty care. Centurion was newly established to bid on this RFP and included as a partner MHM Correctional Services Inc., a firm that held a contract with the MDOC for psychiatrists. In essence, both bidders were insiders to the MDOC system.

Table 1 presents the PPPM figures for Corizon, Centurion, and a system with only public MDOC employees (but with outsourced specialty care and pharmacy). The PPPM figures for Corizon and Centurion came directly from the bid documents. The MDOC figures are for the full public provision of services, based on the staffing plan provided in the RFP (Appendix B) and avoidable costing assumptions, that is, where the RFP required a contractor to absorb an expense, then the item cost shifted to a contractor and was subtracted as an MDOC expense; where the RFP was silent on an MDOC expense, it was assumed the expense would remain a MDOC responsibility.

OPEN IN VIEWER

Table 1.

PPPM Estimates for Bidders and Public Provision.

PPPM item	Corizon	Centurion	MDOC
Staff salaries
Physical health	157.27	137.64	134.29
Mental health	57.02	54.42	60.62
Employee benefits
Physical health	29.08	23.61	68.13
Mental health	8.84	8.41	30.55
Specialty services	87.28	101.85	194.79
Pharmaceuticals
Physical health	45.98	49.73	46.94
Mental health	6.32	4.63	13.93
Indirect costs
Physical health	10.32	19.65	6.16
Mental health	0.49	1.62	2.64
Management fee
Physical health	42.75	21.22	0.00
Mental health	13.20	6.11	0.00
Less savings for both physical and mental health services	4.50	2.83	0.00
Total PPPM	454.05	426.06	557.65

Source. MDOC bid documents and author’s analysis.

Note. PPPM = per prisoner per month; MDOC = Michigan Department of Corrections.

Total PPPM figures at the bottom line in Table 1 show the MDOC cost to be $103.60 higher than Corizon and $131.59 higher than Centurion. It would be wrong, however, to draw conclusions about the relative efficiency of the systems from this one statistic. Observe that the final cost difference between MDOC and rival bidders comes largely from the PPPM estimate for specialty services, an area that has been managed by private contractors since 1998 and was at the time the responsibility of Corizon. The PPPM estimate of $194.79 under MDOC was based on the average MDOC payments to contractors over the 2008 to 2011 periods.

Had the bidders been required to use the same specialty care PPPM figures as the MDOC, the three total PPPM estimates would have been nearly the same. The Corizon estimate becomes $561.56, or $3.91 higher than the MDOC. The Centurion total PPPM becomes $519.00, or $38.65 less than the MDOC. Neither bid would have reduced physical and mental health care costs to the level required by the RFP.

It was revealing that both bidders planned to lower specialty service expenses by about half. The bulk of this expense is for inmate treatments that are unavailable in the prison clinics. Two explanations are possible: first, that the rate of frivolous specialty procedures was exorbitant. If so, then the MDOC is spending more than $50 million annually in excess fees for these services, ⁵ and Corizon is complicit. Second and more troubling is that the bidders planned to reduce the PPPM for specialty care by denying meritorious medical treatments.

What seems apparent is that bidders were not going to reduce specialty care by augmenting treatment in clinical care. Evidence can be found in the proposed staffing: the planned FTE at Corizon and Centurion was 1,372.4 and 1,320.6, respectively. The number of FTE at the MDOC facilities for the equivalent care requirement was 1,550 (RFP, Appendix B). Furthermore, as illustrated in Figures 1 and 2, bidder staffing plans shifted a greater share of the care burden to lower paid occupations.

OPEN IN VIEWER

Figure 1.

Physical health staffing, MDOC, and private bidder proposals.

OPEN IN VIEWER

Figure 2.

Mental health staffing, MDOC, and private bidder proposals.

Figure 1 matches the FTE counts for the MDOC with the private bidders for occupations involved in the provision of physical health care. Across the occupations, the MDOC staffing plan features a higher proportion of personnel with advanced medical skills. The difference is most noticeable with nursing; the MDOC staffing had far more registered nurses (RN), whereas the private bidders were staffed with a greater proportion of licensed practical nurses (LPN) and, to a lesser extent, resident care aides (RCA).

Substituting LPNs for RNs reduced compensation by approximately $10,000 per FTE. Taking such an action, however, might affect operations. By MDOC policy, RNs can administer drug prescriptions, while LPNs cannot. Fewer RNs would mean a greater reliance on physicians for ministering drugs. The MDOC staffing opted for the more expensive RN workforce to free physician time for other activities and to reduce on-call physician services.

A similar pattern was observable with mental health staffing. As illustrated in Figure 2, the private bidders planned to replace positions with less credentialed and lower cost alternatives. Like physical health, the greatest difference is within the middle-range occupations, in this instance psychologists and social workers. The switch was again economically driven; substituting a licensed psychologist with a licensed social worker would yield an approximate annual savings of $14,000 per FTE.

Since most of the value in health care comes from personnel, all things equal, a staffing plan with two hundred fewer FTE or one relying on lower qualified personnel is inferior. One way to improve the precision of the comparison—to compare relative value instead of relative cost—is to overlay the Michigan civil service salary and benefits schedule on the staffing plans by Corizon and Centurion. Table 2 provides a comparison of the PPPM for labor expenses under the private staffing scenarios.

OPEN IN VIEWER

Table 2.

Labor PPPM for Public Provision with Bidder Staffing Plans.

PPPM item	Corizon plan	MDOC with Corizon plan	Centurion plan	MDOC with Centurion plan
Staff salaries
Physical health	157.27	115.92	137.64	101.96
Mental health	57.02	53.42	54.42	55.45
Employee benefits
Physical health	29.08	58.81	23.61	51.72
Mental health	8.84	27.10	8.41	28.13
Total labor PPPM	252.21	255.25	224.08	237.27

Source. MDOC bid documents and author’s analysis.

Note. PPPM = per prisoner per month; MDOC = Michigan Department of Corrections.

As Table 2 indicates, salaries are lower and benefits higher when public civil service rates are imposed on the Corizon and Centurion staffing plans. This occurs in part because people trade-off salary for better benefits (Olson 2002) and in part because the public MDOC compensation system evolved from bilateral negotiations between the state and labor unions, and when workers collectively shape compensation packages, they give substantial weight to benefits. The MDOC benefit levels are based on an estimated roll-up of 50.8%, whereas bid documents indicated that Corizon and Centurion had roll-ups of 18% and 17%, respectively.

After equalizing the staffing plans, differences in labor cost between private and public provision become inconsequential. Total labor PPPM, the bottom line in Table 2, varies by a few dollars. Add the indirect and management fee expenses in Table 1, and privatization is costlier than public provision. The only outstanding item was the expense for specialty services, and neither bid document explained how specialty services were to be reduced so dramatically.

The state-appointed committee evaluating the bid proposals recommended against privatization. The reasoning did not mention the FTE cutbacks and personnel substitutions, but ruled that economic feasibility was not possible because a large share of the MDOC personnel benefits (12.3% of salaries) were accrued obligations and therefore unavoidable. Legacy costs were estimated at $24.4 million for physical health and $7.5 million for mental health. ⁶

Lessons on Prison Health Care Privatization

With the caveat that case studies are prone to idiosyncrasy, the Michigan case offers tentative lessons for public administrators who want to better assess the value of privatization and for interested groups outside of the procurement process that wish to critique privatization proposals. Based on the reasonable assumption that inputs affect the scope and quality of output, the method herein adjusts price by inputs to arrive at relative value. The technique is a refinement to the avoidable cost comparison method that can be used in situations where the principal can require bidders (private for-profit, non-profit, or public entity) to provide a bondable commitment to labor, machinery, and raw material inputs. The approach should be especially useful for complicated public services, where forecasting change in service character due to outsourcing is challenging and where service function quality can have severe human consequences.

The first lesson of the study is that input adjustments can clarify instances where price alone presents a borderline decision. Ambiguity and uncertainty clouds any make-or-buy case this large and complex, creating space for subjectivity and bias. A review committee could have referred to the bottom line of Table 1 and judged privatization as superior. Instead, using an avoidable cost method that carved out accrued liabilities, the MDOC committee ruled otherwise.

Adjustments to cost based on staffing disparities sealed the evidence of public sector superiority. More critical, the bidder staffing plans raised legitimate questions about potential changes in the scope and quality of care. Medical care scope and quality is largely a function of personnel skill and effort, and both bidders submitted proposals to reduce medical FTE and hire staff with fewer qualifications. Clearly, the “efficiencies” to the system were to be achieved through deprofessionalization. And while it might have been possible to move forward without harming inmates, evidence suggests that similar reforms can take place within public systems (Garden et al. 2003).

A second lesson is that privatization ushers in a different approach to personnel management. The MDOC compensation model evolved through bilateral negotiations between the state and public unions, and emphasized components that are important to senior workers, such as health insurance and pensions. The private bidders unilaterally established their compensation plans, which devoted a larger share of the compensation package to salaries. Such differences affect recruitment. By addressing the needs of mature workers, the bilateral approach is geared toward longer-term, stable workforce development. By emphasizing salaries, the bidders structured compensation for the rapid recruitment of younger workers or the retention of experienced state workers, who have the option to retire with state-provided pension and health benefits.

A third lesson concerns the feasibility of competitive bidding as a method of involving private firms in prison health care. Conceptually, competitive bidding admits provider turnover, which disrupts the operational bonds between functional units that are important to custody and security. Discontinuous doctor-patient relationships, moreover, can decrease the accuracy of diagnoses, which may increase unnecessary treatments and drug overprescriptions. In short, the theoretical appeal of managed competition is at odds with the advantages associated with provider stability.

It is telling that the RFP only attracted two bidders, and both had standing formal relations with the MDOC. Embedded firms like Corizon and Centurion have inherent advantages over outside competitors because operational knowledge enhances proposals this large and complex. An absence of genuine competition explains the history of health service privatization at the MDOC, which can be characterized as accommodation toward contractors and a high tolerance for underperformance. Replacing mediocre vendors is costly and disruptive for large and complex services.

A fourth lesson is the unresolvable tension between best practice medical care and the profit motive. The admonition that medical decisions should be made based on “good clinical judgment” (NCCHC 2008, 28) alludes to the challenge. How to achieve medical professionalism when contractual pathways for maximizing profit conflict with best practice? Privatization is usually embarked upon to reduce prison expenses, with efficiencies induced by financially stressing the service function. For medical care, the most direct way to reduce expenses is to deny, delay, or dilute care. Throughout the MDOC history, contractors responded to these pressures by restricting prisoner access to medical care. In this study, the bidder staffing plans were supportive evidence that the contractor response to the combination of fiscal stress and profit motive is to reduce the scope and quality of inmate health care.

Public oversight is therefore critical to ensure that efficiency does not become synonymous with inmate neglect. Contractual covenants alone are insufficient for ensuring responsible contractor behavior because language is effective only if enforced. Monitoring must be robust and conspicuous. The history of health services privatization at the MDOC illustrates how the passage of time can lead to complacency by a public agency and state to the point where contract violations are disregarded and safeguards are dropped. Monitoring health care is costly, in part because doctor-patient interactions are hard to directly observe and in part because monitors must have qualifications that confer enough legitimacy to overrule contractor decisions. Intervention is especially needed at key evaluative points in the system, such as referral to specialty care.

A final lesson is to respect the strength of privatization ideology. Aman (2007) described how a private contract for medical care in NYC resulted in inmate harm and how the response was not to resocialize care but to contract with another provider, which also proved to be negligent. Michigan likewise experimented with private prison health care for two decades before this recent exploration with very little record of success. Instead of changing course, elected leaders pursued more extensive privatization. The remarkable political resiliency of the concept suggests that privatization is not simply an administrative tool plied rationally for the public good. In many contexts, overcoming ideology might require collaborations between social activists and their usual adversaries, conservative political leaders, where it is possible to forge a common goal of achieving fiscally prudent and accountable government.