Privatization
December 15, 2011

Prison privatization risks
higher costs for Ohio
Bob Paynter

Preface to the report
Last spring, when the Ohio General Assembly was considering a proposal by Gov. John Kasich to
sell five state prisons, Policy Matters Ohio released a report by journalist Bob Paynter. The report,
Cells for Sale: Understanding Prison Costs & Savings, examined the cost issues in Ohio’s
privately-run prisons. It found that the cost calculations performed over a number of years by the
state of Ohio had not reliably demonstrated the private-prison savings required under Ohio law. In
the end, the state wound up selling one prison, privatizing the management of two others, and
combining the operations of two more, one of which had been privately run for the last decade.
Now, in the following pages, Paynter has followed up with an analysis of the cost issues with those
new arrangements.
To be clear, Policy Matters believes that there are many reasons to be concerned about prison
privatization and to be skeptical even if cost savings exist. If “savings” result from higher prisonerto-guard ratios, more crowding, less psychiatric or medical care, lower hourly pay for guards, less
spending on training, or less spending on prison education, food, or recreation, they will likely cost
us more in the long run. All of these can result in lower quality supervision, more morale problems,
or overall declines in quality. This in turn can mean less rehabilitation, more recidivism, and
possible increases in tension within the facility.
Second, accountability and transparency are significantly lower in private companies than in public
entities. Incarceration is, by definition, a tremendous incursion on liberty. The isolation and
tarnished reputation of inmates makes them prime targets for abuse. Thus, this is an area of
government where we should want the highest levels of transparency and accountability, even if
that results in slightly higher costs.
Third, private prisons must have room for profit. This can result in perverse incentives, where the
interest of the operator is to maximize the number of inmates. That is not the same interest as those
of Ohioans. In addition, if compensation of guards is lower, then Ohio communities are receiving
less of the return for this state spending, increasing what economists call “leakage”. In other words,
if pay is reduced but profits for out-of–state operators are increased, the money is not staying and
circulating in the Ohio economy. This reduces the positive impact of state spending, making it less
helpful to Ohio’s beleaguered communities.

www.policymattersohio.org

Prison privatization risks higher costs for Ohio

The above concerns are all reasonable to raise about publicly-owned, privately-managed prisons.
The concerns intensify when we consider outright private ownership, as at the Lake Erie
Correctional Institution in Conneaut, which is being sold to Corrections Corporation of America
(CCA). Once prisons are sold or privatized, there is reason to worry that the state will have reduced
bargaining power to change course, giving private operators more leverage and more ability to
garner excessive profits.
For these reasons and more, critics are rightly skeptical about prison privatization and Policy
Matters Ohio opposes it. A demonstration of lower cost is not sufficient reason to give over this
most sensitive government function to private, profit-making companies. This study finds that the
new prison privatization arrangements that the State of Ohio has approved could well turn into a
loser for Ohio taxpayers. And as Paynter points out, the bulk of the savings the state now expects to
achieve with its prison overhaul are to be accomplished not by privatization, but rather by
combining a prison that was previously run by a private operator with a state-run facility under state
control. By the state’s own account, this is expected to save more than twice as much each year as
the private prison operator had supposedly accomplished with one of them.
The sale of the Lake Erie institution has been dubbed by CCA as the first time an American state
has ever sold a prison to a private company. If CCA predictions are correct, more such sales around
the country could follow. This makes it all the more important that we follow this transaction and
how it works out. We encourage you to read this detailed examination of Ohio’s prison overhaul.
– Amy Hanauer,
Executive Director

ii

www.policymattersohio.org

Prison privatization risks higher costs for Ohio

Executive Summary
In the spring of 2011, Ohio Gov. John Kasich proposed the sale of five Ohio prisons – a dramatic
expansion of the state’s 10-year-old prison-privatization experiment – as a way to save millions of
dollars in the face of a deep budget crisis.
By September, after all the numbers had been crunched and the bids received and analyzed, the
scope of the new prison realignment was considerably less dramatic: Only one prison – the Lake
Erie Correctional Institution, in Conneaut, was to be sold; a second – the North Central Correctional
Institution (NCCI), in Marion, was to be combined with a newly reopened former juvenile facility
nearby and operated by a private vendor; and a third – the
Key findings
North Coast Correctional Treatment Facility (NCCTF), in
Lorain County, which had been operated by a private company
• Prison sale could cost millions
for the previous 10 years, was to be returned to state
more than keeping state
supervision and merged with the nearby Grafton Correctional
ownership.
Institution.
• Half of claimed savings
Despite its scaled-back proportions, the state still claimed that
the prison restructuring would save taxpayers about $13
million a year. But just as with earlier claims – that the private
operation of two state prisons had saved Ohioans more than
$45 million from 2000 to 2010 – taxpayers have a right to be
highly skeptical about that figure.

comes from switching prison
back to state operation.
• State’s estimated cost
savings based on flawed
assumptions.

A close look at the sale of the Lake Erie facility to Corrections Corporation of America (CCA)
suggests that that deal, rather than saving up to $3 million a year as the state projects, could easily
wind up costing taxpayers millions of dollars instead. In addition, the state’s claim that private
operation of the combined Marion facility will save another $3 million a year is based on what
appear to be highly dubious accounting assumptions that one expert calls “bogus” and that seem to
bear little relation to reality.
In fact, to the extent that the state’s savings claim is accurate at all, it represents as much a
repudiation of prison privatization as an endorsement. More than half of the purported savings – $7
million a year – would come from converting NCCTF from private to state operation and merging it
with the state-operated Grafton facility, saving by consolidating operations. The state could have
ordered such a move at any time over the last decade. And since the $7 million in annual savings
from that merger is well over twice what the state claimed that NCCTF was saving each year by
virtue of its private operation, what does that move say about the efficacy of prison privatization?
In April, a Policy Matters study raised fundamental questions about the claims of hefty savings
from privatization, concluding that the state’s calculations were not only riddled with errors,
oversights and omissions of significant data, but also were tainted by controversial accounting
assumptions that many experts considered deeply flawed.
At the time, state officials acknowledged that their previous calculations – which were done to
establish that privately operated prisons were meeting Ohio’s legal requirement that they save at
least 5 percent against what it would cost the state to operate them – were both inconsistent and
imprecise. Even as the prison-sale proposal went out for bids, officials were still tinkering with
iii

www.policymattersohio.org

Prison privatization risks higher costs for Ohio

fundamental elements of the calculation. When the new savings formulas were released in May, it
was clear that rather than “updating” the methodology, as the Kasich administration had claimed it
was doing, the Ohio Department of Rehabilitation and Correction (ODRC) had blown up its old
method for calculating purported savings and started from scratch.
For instance, new state spreadsheets showed that a completely new set of comparison institutions –
state-run prisons used as benchmarks to help determine what the costs would be if the state were
running the private facility – had been introduced. The substantial “savings” that privatization
purportedly would produce in ODRC’s central office expenses – a past claim that was based on
dubious accounting assumptions pointed out in the April Policy Matters study – were slashed by
more than 83 percent in the new model (from $3.94 per inmate per day, down to just $0.66). And a
whole new set of supposed cost-saving sources had been rolled out to make up the difference.
Problem is, the rosy savings projections from those sources are based on some of the same flawed
accounting assumptions that tainted the old calculations. Once the likely savings are trimmed to
more realistic proportions, the Marion privatization looks more like a break-even proposition, with
savings, if any, falling well below the 5 percent threshold required by law. And it could turn into an
actual loser for taxpayers.
Meanwhile, based on the terms of the deal and on the numbers available from both parties, the sale
of the Lake Erie facility to CCA – along with the 20-year agreement to have CCA operate it – may
not produce any taxpayer benefits at all. In fact, assuming that all terms remain constant over the
life of the deal, it potentially could cost taxpayers millions of dollars more than if the state had
retained ownership of the prison and allowed the Utah company that currently runs it under contract
to keep operating it. (State officials say taxpayers will save because any capital improvements to the
12-year facility – among the newest of Ohio’s prisons – must now be borne by the new owner. But
they provided no numbers to justify the claim).
One thing you can probably be sure of is that the terms of the deal won’t stay unchanged. It’s likely,
in fact, that CCA will try to extract more favorable terms from the state down the road, which could
increase costs to taxpayers. That’s both because that’s part of its business strategy and because,
without new terms, the Lake Erie facility looks like a less efficient money maker for CCA than its
other prisons.
There is no certainty the company will seek or obtain such a boost. The state maintains that it
doesn’t have to grant such a request, and that any price increases would be capped under the
contract. But CCA will own the prison. And because it will have the right to fill the facility with
out-of-state inmates if talks break down and Ohio is compelled to terminate the deal, an argument
could be made that CCA would have the upper hand in negotiations. Especially with a state that, for
the moment at least, has an overcrowded prison system that needs every bed it can get.
Bottom line? This deal looks like it could turn into a loser for Ohio taxpayers. Only time will tell if
it will.

iv

www.policymattersohio.org

Prison privatization risks higher costs for Ohio
Introduction
On the surface, the state of Ohio’s decision to sell its 12-year-old prison in the city of Conneaut, in
Ashtabula County, to Corrections Corporation of America (CCA) looks like a pretty sweet deal for
taxpayers sweating through some very tough economic times.
At first blush, the sale price looks particularly attractive. The state will get $72.8 million in cash for
a facility that cost $43.9 million to build in 1999 (plus the 120 acres of real estate under and around
it). On the back of the napkin, that looks like a $28.9 million profit for the state. Sweet.
And based on official pronouncements, the terms governing operation and management of the Lake
Erie Correctional Institution over the next 20 years also appear to look good for taxpayers. Not only
does the sale appear to meet the state’s “baseline budget requirement” of generating $50 million in
upfront cash for the Ohio Department of Rehabilitation and Correction. But if you believe the
state’s calculations, the deal with CCA is also projected to save taxpayers nearly $3 million a year
for operations against what it would have cost the state to run the facility. Sweet and sweet.
So, what’s not to like?
Well, you know what they say about the devil and the details. And the details reveal the possibility
of serious and very costly problems with this deal. And not only with this one. A close look at a
second prong of the state’s recent prison realignment decision shows that it, too, may be promising
far more of a bonus for taxpayers than it’s likely to deliver.
In addition to the CCA deal, the state also announced on Sept. 1 that the North Central Correctional
Institution in Marion, combined with a closed juvenile facility soon to be reopened as an adult
prison, would be operated by a private vendor at a supposed savings for taxpayers of nearly $3
million a year. Ohio law requires that any privately run prison realize an operating savings of at
least 5 percent compared to what it would cost the state to run the same facility. Based on the state’s
calculations, a privately operated Marion facility would save about 5.8 percent. But the new cost
calculations the state devised to come up with these savings projections are “bogus,” according to
one seasoned corrections expert, and dramatically inflate the likely savings. Once the projections
are trimmed to more realistic proportions, the Marion privatization looks more like a break-even
proposition, with savings, if any, falling well below the 5 percent threshold.

Lake Erie Correctional Institution
But first, the sale of the Lake Erie facility to CCA. Among the five institutions that Ohio put up for
sale last spring, this is the only one that’s actually being sold. According to CCA, the deal
represents the first-ever sale of a corrections facility by a state to a private company.1 That makes

1

Corrections Corporation of America, “CCA announces 2011 third quarter financial results,” Press release, Nov. 2,
2011, available at http://ir.correctionscorp.com/phoenix.zhtml?c=117983&p=irolnewsArticle&ID=1625258&highlight=
www.policymattersohio.org

Prison privatization risks higher costs for Ohio

Ohio a trailblazer – a CCA executive’s term2 – among privatization advocates. But was it a good
deal for taxpayers?
Focusing strictly on dollars and cents, the truest test of whether the CCA prison-sale agreement is
boon or bust for Ohio is simply this: Going forward, will it wind up costing taxpayers less (sweet)
or more (not so sweet) than if state officials had either left well enough alone or gone in a different
direction?
Since the deal includes an agreement that CCA will operate the facility for 20 years, some
assumptions about the future must be made. But after adopting even the simplest, most basic (and
least probable) of assumptions – that the terms will remain constant over the life of the agreement –
a closer look shows that this deal has the potential to be a net loser for taxpayers right off the bat.
And as more probable circumstances are contemplated, the potential downside for taxpayers starts
to grow, producing plausible scenarios under which the state could be paying millions of dollars
more each year to keep the Lake Erie facility running under the CCA deal than if it had simply
maintained the status quo.
First, the simplest scenario: Assume that the announced terms remain constant over the life of the
agreement with CCA: The $72.8 million purchase price is paid to the state in cash upfront; the
population at Lake Erie stays at 1,798 inmates throughout, and the agreed-upon “per diem” – the
amount the state pays to CCA per inmate per day to run the facility – holds steady at $44.25 (The
agreement guarantees per-diem payments for at least 1,618 prisoners, or 90 percent of the facility’s
capacity). The “annual ownership fee,” the amount the state has agreed to pay CCA each year for
the use of the facility (now owned by CCA) to house Ohio prisoners, also stays at $3.8 million.
Projected over the 20-year life of the agreement, those terms would cost Ohio taxpayers a net of
anywhere from $8 million to $15 million more than if the state had simply decided to stick with the
current private operator at Lake Erie and to accept its new bid.
Where in this wide range the number would actually fall depends on what Ohio’s financing costs
were on the bonds it issued to pay for the original prison construction in 2000. Officials at the state
Office of Budget & Management say they can’t provide those figures because multiple bond issues
were involved, at multiple rates and terms. If Ohio had to pay 7.5 percent on a 20-year issue, for
instance, well above the going rate at the time, the net cost to taxpayers over the life of the CCA
deal would come in at around $8.2 million. If Ohio paid the likelier interest rate of about 5.3 percent
– roughly the municipal bond rate for 20-year notes issued in 1999 – the net cost to taxpayers would
be about $11.4 million as Table 1 shows. At 2 percent – which assumes a substantial interest
savings for Ohio by the mixing and matching of bond issues – the net cost over the 20 years would
come in at $15.1 million.

2

2

Corrections Corporation Of America Third Quarter 2011 Earnings Conference Call, Nov. 3, 2011
www.policymattersohio.org

Prison privatization risks higher costs for Ohio

Table 1

Cost implications of sale to CCA vs. operation by MTC
Costs to taxpayers
Annual ownership fee (AOF)
Operations per prisoner, per diem
Lake Erie inmate population
Operating costs per year1
Operating costs over 20 years
Ohio’s likely debt-retirement costs2
Costs
Income from purchase price
Total cost over 20 years3
Net cost to taxpayers

CCA
$3,800,000
$44.25
1,798
$32,839,948
$656,798,950
$23,211,657
$680,010,607
$72,770,260
$607,240,347
$11,398,809

M TC
-$43.22
1,798
$28,363,989
$567,279,788
$28,561,750
$595,841,538
-$595,841,538
--

Source: Author's analysis of data from the Ohio Department of Rehabilitation and Correction. Calculations
assume simplest scenario, that all terms stay constant for 20-year deal. CCA is Corrections Corporation of
1
America; MTC is Management and Training Corporation. Notes: Calculated as per diem x population x 365, plus
2
AOF; Assuming initial bond rates of 5.33 percent for 20 years, these would be Ohio’s costs to pay outstanding
3
principal after CCA sale or, if no sale (MTC), to pay original loan to maturity; Costs minus revenue from
purchase.

When asked to explain, state officials argued that the sale to CCA will prove to be the “most
financially advantageous” option available to taxpayers because it will allow the state to avoid
future capital-improvement costs for the 12-year-old Lake Erie prison, which will shift to the new
owner – and that “major repairs are upcoming.” That may be true. But they provided no figures to
justify the claim, making its impact on costs difficult to evaluate. It should be noted that in
November, shortly after CCA agreed to the purchase, the company announced that it would sink
$3.1 million in capital improvements into the Lake Erie facility, one of the state’s newest.
Assuming that’s an expenditure Ohio avoided, it still doesn’t make up for the excess costs taxpayers
appear to be taking on with this deal.
Had the state simply stuck with its current arrangement – contracting with Utah-based Management
& Training Corporation (MTC) to run the Lake Erie facility – the state also would have retained
ownership of an asset worth up to $73 million, free and clear.
Documents show that among the proposals submitted to the state only for operation and
management of the Lake Erie facility (with no sale), MTC was the low bidder, proposing a per diem
rate of $43.22, compared to CCA’s original bid of $45.86. It was only when purchase proposals
were thrown in that CCA became the low bidder. (Its final per diem was trimmed to $44.25 in
subsequent negotiations with the state.)3
But as “real world” assumptions about the agreement are contemplated, net cost projections for
Ohio taxpayers could easily get worse. Other than the purchase price, which is a fact on the ground,
3

Ohio Department of Rehabilitation and Correction document: Operation, Management and Purchase of Correctional
Facilities Cost Scoring Evaluation – DRC001/CSP901412 – August 24, 2011
3

www.policymattersohio.org

Prison privatization risks higher costs for Ohio

and maybe the inmate population at Lake Erie (which already represents a 300-bed increase over its
current population to bring it more in line with 30 percent overcrowding system-wide) – very little
about this deal should be expected to remain constant over 20 years.
Both of the most critical financial terms – the operation and management (O&M) per diem and the
annual ownership fee (AOF) – will be subject to re-adjustment every two years, as the state
considers and approves its new budgets. Because of the state law requiring private O&M costs to
stay at least 5 percent below what it would cost the state to run the facility, the AOF would arguably
be the more flexible figure. And it should shock no one if the company tries to jack that figure up in
future budget cycles.
Because CCA will own the Lake Erie facility, language in the deal suggests that the company could
enjoy the upper hand in future fee negotiations, unless the state succeeds in dramatically reducing
its prison population. Under the agreement, if problems develop between the two parties and the
state is compelled to terminate the O&M and AOF portions of the contract, CCA would have the
option to fill the facility with out-of-state prisoners at a potentially higher daily rate. The state, in
turn, would have to move its Lake Erie prisoners into what are now overcrowded facilities
elsewhere in Ohio or go shopping for available prison beds in other states. In either case, it seems
possible that net costs for Ohio taxpayers from this deal could rise as the agreement matures.
There are at least two reasons to expect future CCA efforts to hike fees. First: As company officials
have made quite clear in their public filings, that’s what they do. (Part of its three-pronged strategy
for growing its business, CCA reports, is “enhancing the terms of our existing contracts.” 4) And
second: Judging from its public filings and comments, the purchase of the Lake Erie Correctional
Institution doesn’t look like a very good deal for the company unless it DOES manage to boost its
fees.
In making a substantial capital investment involving some risk – the purchase, for instance, of a
1,798-bed prison in northern Ohio – most prudent investors would be looking for an annual return
(operating revenues, less operating costs, as a share of investment) of at least 10 to 12 percent. CCA
officials told stock analysts in early November that they normally shoot for a return of 13 to 15
percent on capital investments5 and expect that their Ohio prison purchase will meet that target. But
after taking all of the revenue (O&M per diem and AOF) that the company apparently will receive
from Ohio under the Lake Erie agreement (assuming no “enhancements” to the financial terms), and
subtracting its likely costs, the deal stands to generate a return of about $4.96 million a year for
CCA.6 Measured against its purchase price of $72.8 million, that looks like an annual return of just
6.8 percent – well below what a prudent investor would have anticipated and what CCA’s
stockholders have come to expect. If the company borrows, say, half the purchase price and
measures its return only against the cash it has in the deal, than it would appear to meet the 13
4

CCA’s 2010 Annual Report on Form 10-K, page 10
Before interest, taxes, depreciation and amortization. Corrections Corp. of America Third Quarter 2011 Earnings
Conference Call, Nov. 3, 2011
6
Likely costs are based on the actual costs per inmate per day ($42.48) that the company says it incurred in 2010 in the
45 facilities that, like Lake Erie, it both owns and operates. About 93 percent of the 64,000 beds in those facilities are
for prisoners, like those at Lake Erie, who are at a medium security level or lower. Information from CCA’s 2010
Annual Report on Form 10-K, pages 42 and pages 5-7.
5

4

www.policymattersohio.org

Prison privatization risks higher costs for Ohio

percent threshold. But it still would have acquired what appears to be an underperforming asset.
CCA’s expected annual return at the Lake Erie facility adds up to a gross profit margin that’s just
over 40 percent of what the company typically produces at its owned and operated facilities
elsewhere. (See Table 2)
Table 2

Lake Erie Correctional Institution: An underperforming asset?
CCA’s purchase price: $72,770,260

CCA’s Lake Erie
operating revenues
Anticipated
operating costs2

Per diem

Population

Annual
operations1

Annual ownership
fee (AOF)

$44.25

1,798

$29,039,948

$3,800,000

$32,839,948

$42.48

1,798

$27,878,350

--

$27,878,350

Return (annual revenues minus annual costs)
Return as percent of purchase price

Annual
totals

$4,961,598
6.8%

Comparison of gross profit margins
Lake Erie per diem3
Operating revenues
Operating costs

2

Operating profit
Gross profit margin
(% of revenues)

All CCA’s owned and managed
facilities 2010 per diem

$50.04
$42.48
$7.56

$66.30
$42.48
$23.82

15.1%

35.9%

Source: Author's analysis of data from Corrections Corporation of America and the Ohio Department of Rehabilitation and
1
2
Correction. Notes: Calculated as per diem x population x 365; CCA’s average cost per prisoner per day for its 45 other
owned and managed (O&M) facilities nationwide, totaling 64,000 beds, practically all of them (93%), like Lake Erie,
3
medium-security or lower; Operations per diem, combined with AOF, expressed as per prisoner/per day.

Those numbers raise four distinct possible scenarios:
1. That there are more revenues for CCA in this deal – besides the O&M per diem and the
AOF – than meet the eye in a review of the purchase and operating agreement. (State
officials say there are no additional revenue sources.)
2. That CCA’s operating costs for Lake Erie will be substantially lower (roughly 20 percent
lower) than the average at their 45 other owned and managed facilities, raising questions
about what level of services Ohio will actually be getting for its money. (State officials say
CCA will be providing full-service operation and management at Lake Erie.)
3. That CCA and its stockholders, for whatever reason, are willing to settle for an Ohio asset
that appears to be performing at well below the levels they’ve grown accustomed to.
(Despite repeated phone messages and emails, CCA officials declined to respond to specific
questions about their return-on-investment projections or about cost- and revenueexpectations for the Ohio facility.)
4. That CCA will be pressing Ohio for fee “enhancements” to bring Lake Erie’s performance
more in line with corporate expectations.

5

www.policymattersohio.org

Prison privatization risks higher costs for Ohio

As already noted, Ohio taxpayers could well be bleeding red on the deal even if no changes are
made. Eliminating scenarios 1, 2 and 3 for the sake of argument, if CCA succeeds in enhancing its
fee to the point where Lake Erie’s operating profit starts to approach company averages elsewhere,
the hemorrhaging could accelerate dramatically.
(ODRC officials say any dramatic fee hikes for CCA are impossible because of wording in the
contract, but an exception to that wording appears to call that position into question.)7
Assume, as the state avers, that there are no hidden revenues and that the O&M per diem stays
about the same (recalling that the O&M per diem is fairly inflexible because of Ohio’s 5 percent
savings law). Assume further that CCA’s costs per Ohio prisoner are comparable to what it spends
on medium-security prisoners elsewhere. CCA would need to more than double the annual
ownership fee – from $3.8 million to $8.3 million – to realize a return of 13 percent, if return is
measured against its purchase price. Or, in the alternative, it would need an AOF of more than $11
million to push its gross operating margins beyond 30 percent (still considerably shy of the 35.9
percent average it enjoys for similar facilities elsewhere). Over the life of the agreement, such an
adjustment would bring the net cost for taxpayers for the CCA deal over the status quo to as much
as $8 million. Every year. For the rest of the 20-year agreement. And at the end, unlike with the
status quo, Ohio would no longer own the prison.
Now, there is no way of knowing if CCA will ever seek fee hikes of anything approaching that size.
You might argue, of course, that even if it did, Ohio officials could just say “no.” And that’s true.
ODRC need not propose, and the General Assembly need not approve anything beyond modest,
inflationary increases in the O&M per diem or the AOF.
But to predict that as a successful cost-containment strategy is to predict that CCA will accept what
appears to be anemic production from its Ohio property ad infinitum – even though it might be able
to generate as much as 33 percent more in per-prisoner revenues at Lake Erie if it were able to
obtain prisoners from elsewhere, which would leave Ohio in something of a lurch.
(It should be noted that in 2008, CCA threatened to replace all Colorado inmates in one of four
facilities the company owns there with prisoners from another state that paid more unless Colorado
granted its demand for a 5 percent increase in its per-diem rate. The threat prompted a key
lawmaker to declare that CCA had a “disproportionate amount of negotiating power.” 8 As a
compromise, the state legislature approved a 4.25 percent increase for CCA in March of that year.
As state revenue projections plummeted, however, the legislature rescinded the rate increases for
CCA and all other private service providers in February 2009.)
Under the agreement between Ohio and CCA, if the state terminates the operating and management
portion of the agreement for whatever reason (perhaps because CCA can’t do a satisfactory job for
the money Ohio agrees to give it), then Ohio can bring in another private entity to run the facility.
7

State officials deny the possibility of sizeable hikes in CCA’s fees, citing language in section 9.3 of the contract that
limits increases to standard measures of inflation in wages and costs. But that same section of the contract clearly makes
an exception to those limits “for any changes in compensation negotiated as a result of an amendment to this contract.”
It should be noted that contract amendments, many of them dealing with compensation and other financial matters, have
been routine in the state’s relationship with MTC, the current private operator of the Lake Erie facility. State records
show at least eight such contract amendments in 10 years.
8
Ed Sealover, “Private Prison Wants More Money,” The Colorado Springs Gazette, Feb. 10, 2008.
6

www.policymattersohio.org

Prison privatization risks higher costs for Ohio

Or ODRC could choose to run it itself. But since CCA will own the facility, the state would still
have to pay an annual ownership fee in order to use it for Ohio’s prisoners. According to the
contract, that fee – the AOF – would have to be renegotiated to reflect the new, post-O&M
relationship. ODRC officials say the state can terminate the O&M portion of the contract and keep
the AOF exactly as it is, without negotiating with CCA unless the state wants to – a position that
appears to be directly contradicted by language in the contract.9 Depending on how it calculates its
return on investment, CCA would still need to get millions of dollars from Ohio taxpayers each year
in ownership fees in order to realize its baseline return of 13 percent. That’s in addition to whatever
amount taxpayers would have to spend to have ODRC run the facility itself or hire another
contractor do it.
What would happen, however, if the negotiations over a new AOF broke down – if CCA demanded
more money for the right to use its facility than Ohio wanted to pay? As with any other negotiation,
Ohio could either come to terms with the company or terminate the AOF agreement. But that would
also terminate the state’s right to use CCA’s prison. Under the contract, CCA could then try to fill
Lake Erie Correctional with inmates from out of state, and Ohio would have to move its prisoners to
other facilities. 10
For its part, CCA has made no secret that it could follow that course should the circumstance arise.
If the O&M portion of the contract is terminated, and “if it is negotiated that ODRC will not
continue to use the Institution,” CCA noted in its Ohio bid proposal that it would be free to use the
facility to house out-of-state inmates and would in that case “develop a specific transition plan for
the scheduled transfer of ODRC inmates to other State institutions.”11
According to its SEC filings, the typical rate CCA received in 2010 per inmate per day in facilities
it owns and operates was $66.30.12 Compared to the $50.04 it will be getting from Ohio under the
Lake Erie deal (O&M per diem, plus the AOF divided per prisoner per day), that translates to an
annual per-prisoner revenue hike of 32.5 percent.
Then, assuming that Ohio’s prison population has not fallen dramatically, the state would have to
either shoehorn as many as 1,798 Lake Erie prisoners into now overcrowded facilities elsewhere or
ship them out of state at a daily rate – assuming that CCA’s per diems reflect the market – that
could be far higher than what it’s paying now. Or, build another prison to replace the one it sold to
CCA. None of these scenarios appear even remotely more beneficial to Ohio taxpayers than simply
leaving well enough alone in Conneaut.

9

In Section 9.4 of the contract, the language appears clear: “If the state terminates the O&M portion of the Contract, the
AOF will be re-negotiated…” (emphasis added).
10
(Under Ohio law, CCA would have to negotiate a contract with the city of Conneaut authorizing it to take out of state
prisoners. No discussions about such a contract have occurred, but Conneaut City manager Tim Eggleston said in an
Oct. 17 interview that he is certain that city officials would much prefer having out-of-state prisoners at Lake Erie than
no prisoners, and no jobs.)
11
CCA’s Technical Proposal for the Operation, Management and Purchase, Lake Erie Correctional Complex, July 1,
2011, 1 p.m., Section 8, Workplan Lake Erie, p. 103-104.
12
CCA’s 2010 Annual Report on Form 10-K, page 42.
7

www.policymattersohio.org

Prison privatization risks higher costs for Ohio

But the sale of the Lake Erie facility isn’t the only apparent problem with Ohio’s ongoing flirtation
with prison privatization.

Flawed assumptions
In April, Policy Matters Ohio released a study challenging the State of Ohio’s claims of substantial
savings from the operation of two private prisons over the last decade and its use of those purported
savings to justify a proposal to dramatically expand the privatization experiment. A detailed
analysis of the state’s cost calculations showed them not only to be riddled with errors, oversights
and omissions of significant data, but also tainted by controversial accounting assumptions that
many experts considered deeply flawed.13
While criticizing the study, the Ohio Department of Rehabilitation and Correction asserted through
spokesman Carlo LoParo that officials had “updated the private prison cost savings models utilized
by previous administrations … to a model the Kasich administration believes is more accurate.”
LoParo then released a whole new set of numbers – claiming dramatic private-prison savings in the
past and promising more in the future under Gov. John Kasich’s plan to sell off as many as six
facilities – with no documentation of the new “model” or how it had produced these new savings
figures.
When repeatedly pressed for details by Policy Matters Ohio, ODRC produced a set of formulas and
spreadsheets that raised more questions than they answered, while making one point perfectly clear:
The methodologies or “models” that Ohio used in the past to justify the private-prison experiment
and to claim healthy cost savings were essentially worthless – apparently even to the state. Rather
than “updating” those models, as the new administration claimed it was doing, it appears instead to
have blown them up at the 11th hour and started from scratch. ODRC officials said they were still
tinkering with a new formula as late as March 14, a day before the new privatization proposal was
announced and three weeks before the state issued its request for proposals from private-prison
operators. Parts of the new formula were contained in a spreadsheet the department released to a
news organization, in response to a public records request, a month after the bid proposals went
out.14 Another set of spreadsheets containing the rest of the changes was supplied to Policy Matters
Ohio three weeks after that, following another records request.
The new spreadsheets showed that a completely new set of comparison institutions – state-run
prisons used as benchmarks to help determine what the costs would be if the state were running the
private facility – had been introduced. And the substantial “savings” that privatization purportedly
would produce in ODRC’s central office expenses – a past claim that was based on highly dubious
accounting assumptions pointed out in the Policy Matters study – were slashed by more than 83
percent in the new model (from $3.94 per inmate per day, down to just $0.66). In addition, a whole
new set of supposed cost-saving sources had been rolled out to make up the difference.
To get the savings calculation safely over the 5 percent legal hump, the new model introduces new
sources of significant cost savings that were never mentioned before in the state’s private-prison
13

Paynter, Bob, Cells for Sale: Understanding Prison Costs & Savings, Policy Matters Ohio, April 2011, available at
http://www.policymattersohio.org/CellsForSale.htm
14
“FOP Joins Union, Lundy, Policy Matters Against Prison Privatization; DRC Debunks ‘Methodology’ Claim,” The
Hannah Report; Vol. 129, No. 87; May 2, 2011
8

www.policymattersohio.org

Prison privatization risks higher costs for Ohio

analyses and that are based on the same kinds of dubious accounting assumptions that distorted the
discarded central-office “savings” in the first place.
These new “savings” centers formed the backbone of the state’s argument for expanding its privateprisons experiment: the maximum allowable bid amounts for the operation of Ohio’s stable of
private prisons (the highest per-diem amount the state could pay a private operator and still meet the
5 percent savings requirement) were calculated based on those projected savings being realized in
full. And yet, they are rooted in assumptions that are just as likely to prove illusory as not – in fact,
probably more likely. Without the purported savings in these areas being fully realized, the savings
from private operations would fall short – perhaps even well short – of the 5 percent requirement,
removing any legal justification for the privatization experiment.
As it happened, the state dramatically scaled back its privatization agenda after the bids came in.
And the new prison arrangement announced in September appears to be as much a repudiation of
privatization as an endorsement – with the lion’s share of the announced savings actually resulting
from a move away from, not toward, private operation.
According to state documents, no bids were submitted simply to run the Grafton Correctional
Complex, which represents a merger of the Grafton Correctional Institution (GCI), currently
operated by the state, and the nearby North Coast Correctional Treatment Facility (NCCTF), which
has been operated by a private vendor since 2000. And the bids to purchase and operate the
combined facility were way high. (By way of illustration, their cost scores as calculated by the state
were at least 48 percent higher than that of the eventual winning bid to buy and run Lake Erie.)
So, instead of selling the Grafton complex, the state opted instead to return NCCTF to state
operation and merge it with GCI, vowing to save more than $7 million a year “through the
consolidation of administrative operations and costs.”
Ironically, ODRC could have ordered such a move at any time since NCCTF was opened in 2000.
And since the $7 million in annual savings from that merger is well over twice what NCCTF was
supposedly saving each year by virtue of its private operation according to the state’s calculations,
why wasn’t it done years ago? And what does the move say about the efficacy of prison
privatization?
Meanwhile, the supposed “savings” from the rest of the initiative announced in September are far
more dubious, at best.

9

www.policymattersohio.org

Prison privatization risks higher costs for Ohio

North Central Correctional Complex
In addition to the CCA sale, the possible impact of which has already been discussed, and the
merger of NCCTF and GCI, the third leg of the new Ohio prison-restructuring initiative involves
the North Central Correctional Complex in Marion. The complex results from the merger of the
North Central Correctional Institution with a soon-to-be-reopened former juvenile facility nearby.
The state offered this complex for sale as well. But the two purchase bids it received were way
higher still (both scores were more than 83 percent higher than that of the final Lake Erie bid), so
state officials opted not to sell the facility. But they did accept the only bid they received to operate
and manage it – from MTC, the same Utah firm that has been operating Ohio’s only two private
prisons for the last decade.
Measured against the 5 percent savings requirement in state law, however, the margin for error in
the MTC bid is razor thin. In fact, its bid of $41.20 per inmate per day is exactly 1 penny lower than
the actual average costs – before the newly calculated “savings” sources are thrown in – at the three
state-run institutions chosen by ODRC as benchmarks for the facility. In other words, based on
actual measurable costs, the choice between operating the facility publicly or privately appears to be
a wash. (Actually, private operation likely would cost taxpayers slightly more because of the costs
of contract-monitoring services that ODRC is required to provide at private facilities.)
It’s not until the state inserts into the formula the dubious new “savings” sources it derived this
spring that the Marion arrangement appears to reach the 5 percent savings threshold, and even then,
it just barely does. According to the new state formula, if all of these projected savings are realized
100 percent, the MTC bid will produce an annual savings of 5.8 percent.
But after reviewing the state’s method for identifying these new savings sources, Dr. Gerry Gaes,
former research director at the federal Bureau of Prisons and visiting scientist at the National
Institute of Justice, said that any “savings” from the new sources are likely to be much closer to zero
than to 100 percent of the projections. That’s because the state’s method for calculating the
projections are “bogus,” he said15.
Here’s a closer look.
When ODRC revealed its new cost calculations in May, department officials included several new
costs purportedly associated with the running of public prisons that the department had not cited in
previous calculations – costs that would not be incurred for inmates in private prisons and, as a
result, would represent a “savings” of privatization.
The largest among these new “savings” sources was labeled in the new state formula only as
“Statewide.” When pressed for details, state officials described it as “bulk purchases that are shared
by all institutions.” When pressed further (purchases of what?), ODRC supplied a list of about 25
cost items that have varied widely in size over the last three years with almost all of the costs
concentrated in three areas: miscellaneous medical and health-care services for inmates; bulk food
purchases and post-secondary education for inmates. The expenditures, ODRC argued, are made

15

Interview with Dr. Gerald Gaes, director of research, Federal Bureau of Prisons from 1988 to 2002; visiting scientist
at National Institute of Justice, 2002 to 2007. Conducted September 14, 2011.
10

www.policymattersohio.org

Prison privatization risks higher costs for Ohio

only for inmates under the state’s supervision. So, when privatization reduces the number of statesupervised inmates, those “statewide” expenditures should go down as well – ergo, a “savings.”
On paper, that may be a nifty piece of accounting. But in order for there to be a “savings,” the
spending of actual cash money would have to go down – in reality. And there is no way to
determine from the state’s new model how likely that is to happen, or how big such a “savings”
would be. First of all, there has been no consistency in those expenditures from year to year (in
2008, they totaled $5.7 million; in 2009, more than $35 million), making predictions problematic.
So, ODRC simply took the 2010 number ($30.3 million) and declared that to be the “statewide”
expenditure figure for future calculations. But the bigger problem is that the state assumes that all
state-supervised prisoners account for an equal portion of those costs, even though nearly 60 percent
of these “statewide” expenditures were for health care and related expenses and ODRC has said that
only healthier prisoners – arguably less likely to access such services than their sicker counterparts
– are to be assigned to privately run facilities.
The better calculation strategy would have been to use only such “statewide” goods and services
consumed at the benchmark institutions – those considered to be comparable to the private facilities.
Somewhat incredibly, however, ODRC insists that such costs “have not historically been recorded
by inmate or institution.”
In other words, the state has no idea who is accounting for those costs. So, it assumes in its new
savings model that all inmates account for them equally. If the number of state-supervised inmates
declines by 5 percent as a result of the new privatization agreements – roughly the current
projection – then the state’s model assumes that those “statewide” expenditures would decline by 5
percent as well. If, in reality, the actual expenditures were to fall by just half that much instead, then
projected “savings” at the Marion facility would drop to just 3.9 percent, based on the state’s
calculations, or to just 2 percent if those “statewide” expenditures don’t actually fall at all. Both
figures fall well below the 5 percent savings requirement. In addition, both figures further assume
that projections for the other new “savings” centers prove accurate and true. And they are shakier
still.
The problem, Gaes said, is that especially with a system as large as Ohio’s – with roughly 50,000
inmates behind bars – changing the status of even a few thousand prisoners is not going to affect
costs on anything like a pro-rata basis because of economies of scale. That’s not how reality works.
“To make this concrete,’ Gaes said, “if you paid a thousand dollars for mental health and medical
services, and you had 10 inmates, the average cost would be $100 per inmate. But if you went down
to nine inmates, you’re not going to save $100 dollars.” Rather, Gaes said, you’re likely to save
some fraction of that. The same holds true with other functions.
“Let’s say they have 30 people in headquarters that monitor payroll, do human resources, etc. for
the public employees. Now that they will have several thousand fewer inmates and however many
fewer (prison) staff, will they in fact reduce the number of personnel working in that office? My
guess is they probably won’t.” And if they do, Gaes said, the reduction is not likely to be even close
to proportional to the reduction of inmates.

11

www.policymattersohio.org

Prison privatization risks higher costs for Ohio

“To me, the only sensible way of doing this is to go function by function and try to figure out to
what extent, as a result of privatization, have costs for that function been reduced,” Gaes said. “And
typically they aren’t. Or they aren’t reduced dramatically.”
The second largest item among the new savings sources in the state’s model is labeled as OSUMC.
It involves ODRC payments to the Ohio State University Medical Center for outpatient
hospitalization and “telemedicine” consultations provided under contract to the state for inmates
under its supervision. Those services, which cost more than $12 million last year, are not available
under the ODRC contract to inmates in private prisons.
But once again, even more incredibly, state officials insist that “OSUMC costs have not historically
been recorded by inmate or institution.” So, assuming that’s true, then in the absence of any
accurate idea of who is actually using those services, the state has simply assumed that all inmates
are using them equally and that the privatization of 5 percent of inmates will result in a 5 percent
reduction in outpatient and telemedicine charges by OSUMC. (To its credit, ODRC did not try to
argue that the far greater costs for inpatient hospitalization at OSUMC – almost certainly the near
exclusive province of the sickest inmates and those least likely to be put in private prisons – would
be saved in like manner. Inmates of this “higher medical acuity … would be unlikely to assigned to
… a privately operated facility,”16 according to ODRC, or to the benchmark institutions it is
compared to.)
But doesn’t it follow that the healthier prisoners – those most likely to be assigned to a private
facility – would use outpatient OSUMC services less than their sicker counterparts as well? And
wouldn’t the “savings” from privatization in this area be considerably less than the system-wide
average cost for all inmates, sick and healthy alike? If both “statewide” and OSUMC savings come
in at only half the projections, then the privatization savings at Marion top out at about 3.2 percent.
The new state formula makes exactly the same argument with what’s left of the central office cost
savings – that all state-supervised inmates drive those costs equally, and that adding or deleting one
inmate adds or deletes his pro-rata share of ODRC’s costs for such functions as human resources,
payroll, personnel, labor relations, administration, budgeting, accounts payable, the Corrections
Training Academy and about 18 other expense categories.
That may look fine in a spreadsheet calculation. But if you don’t actually spend fewer real dollars as
a result of privatization, then you’ve saved nothing. And in reality, said Gaes, “if you have 10 fewer
prisoners, you don’t pay the average-per-prisoner times 10 less. You have some kind of marginal
decrease.” If all three of state’s purported savings sources max out at 50 percent of projections,
then the savings at Marion would be about 2.4 percent. If costs in those three areas don’t actually
decline at all, then the private Marion facility becomes slightly more expensive to operate (1.2
percent more expensive) than a state-run comparable, according to the new formula. Table 3 shows
how the figures work out in each of these three cases – with all projected savings realized, 50
percent of projected savings realized or no projected savings realized.

16

Email from Austin Stout, assistant chief legal counsel, Ohio Department of Rehabilitation and Correction, sent to Bob
Paynter on June 16, 2001, in reply to public information request.
12

www.policymattersohio.org

Prison privatization risks higher costs for Ohio

Table 3

Ohio’s “savings” projections for North Central Correctional Institution in
Marion, to be operated by Management & Training Corp. (MTC)
Projected facility population: 2,706
Per
prisoner/
per day
Actual expenses if
operated by state1
Projections for newly
added “savings”
centers
Operation support
center2
Statewide (bulk
purchases)
OSUMC (outpatient
medical services)
Total projected cost for
state operation
MTC bid
State contractmonitoring cost
Costs to state under
MTC
Dollar savings (cost)
Percent savings (cost)

Annual
savings/
costs

Per
prisoner/
per day

Annual
savings/
costs

Per
prisoner/
per day

Annual
Savings/
costs

$41.21
100% of savings realized

50% of savings realized

$0.66

$0.33

$1.70

$0.85

$0.69

$0.35

$44.26

$43,715,159

$42.74

No projected savings realized
$0.00
$0.00
$0.00

$42,208,932

$42.21

$41.20

$41.20

$41.20

$0.50

$0.50

$0.50

$40,702,705

$41.70

$41,186,673

$41.70

$41,186,673

$41.70

$41,186,673

$2.56
5.8%

$2,528,486

$1.04
2.4%

$1,022,259

($0.49)
(1.2%)

($483,968)

Source: Author's analysis of data from the Ohio Department of Rehabilitation and Correction.
Notes: 1Average of actual costs for payroll, contracts, supplies and equipment at three state-run prisons considered “comparable” to the Marion
facility; 2The state’s revised “central office” savings.

According to Gaes, the zero-savings scenario is much more likely than the one projected by the
state. There may be some savings in one or all of those areas, he said, but they would likely be
minimal – much closer to 0 than to 100 percent of the per-prisoner average.
“It’s definitely closer to nothing than to the pro-rata share,” he said. “I didn’t say it is nothing, but
it’s definitely closer to it. You’re dealing with the lower end of the savings spectrum.”

13

www.policymattersohio.org

Prison privatization risks higher costs for Ohio

Author
An investigative reporter and editor for nearly 25 years in Akron and Cleveland, Bob Paynter
took an early retirement buyout from The Plain Dealer in 2008. He continues to investigate,
research and write about matters affecting public policy on a contract basis. In April 2011, he
researched and wrote a report for Policy Matters about prison privatization entitled “Cells for
Sale: Understanding Prison Costs and Savings.”

Policy Matters Ohio is a non-profit, non-partisan research institute dedicated to researching an
economy that works for all in Ohio. Policy Matters seeks to broaden the debate about economic
policy by providing research on issues that matter to Ohio’s working people and their families.
Areas of inquiry for Policy Matters include work, wages, and benefits; education; economic
development; energy policy; and tax and budget policy. Generous funding comes from the
Cleveland, Ford, George Gund, Joyce, Sisters of Charity, St. Luke’s, and Raymond John Wean
foundations, as well as the Center on Budget and Policy Priorities, Greater Cleveland
Community Shares and the Economic Policy Institute. To those who want a more prosperous,
equitable, sustainable and inclusive Ohio… Policy Matters.
3631 Perkins Avenue, Suite 4C-E • Cleveland, Ohio 44114
85 E. Gay Street, Suite 802 • Columbus, Ohio 43215

© 2011 Policy Matters Ohio. Permission to reproduce this report is granted provided that credit is given to Policy
Matters Ohio. All rights reserved.

14

www.policymattersohio.org