Proponents of opening the empty jail unit claim that it would "make money" and now have offered a per bed cost projection that is the first attempt by anyone but myself to present any kind of in-depth analysis of the per unit daily cost of owning and operating the Sauk County Jail. Ironically, this new analysis doesn’t evaluate the empty jail unit. The budget also lumps unit B and the Huber unit together in one analysis. For decision making purposes, it is more useful to budget each unit separately.

In the past, I’ve often been told that such an analysis was done by the jail consultant. If anyone else found it in the jail consultants report, they should tell me what page it is on.

Those of us who seek value for the dollars we spend, find that this new per bed cost analysis is overly optimistic, and if used without other information, could easily fool decision makers into very costly mistakes. No single measure provides enough guidance for good decisions about an issue as complex as this one. The analysis would be far more useful if it included a daily cost per Sauk County inmate calculation for all cost factors along with other measures as I did in my analyses.

I say this for several reasons:

1. The most important of these reasons is that the analysis divides the net cost of the operating units of the jail (Huber and Unit B) by all beds in those units instead of by the number of Sauk County inmates.
2. As it is presented, it assumes an 80 year depreciation period to represent part of the annual ownership cost, which is understated partly as a result.
3. Only about 36% of the cost of utilities, routine maintenance and debt service shown on page 376 of the 2005 Sauk County budget book or page 134 of the proposed 2005 Sauk County budget book for the law-enforcement center is included as an annual cost in the analysis.
4. The analysis fails to anticipate cost increases. Employee compensation rates have increased by about 5% compounded annually.
5. Not only does the analysis understate the jail’s share of debt service utilities and routine maintenance, it doesn’t anticipate more major repairs.

Each of these points are described in more detail as follows:

1. The most important of these reasons is that the analysis divides the cost of the operating units of the jail (Huber and unit B) by all beds or cells in those units instead of inmates. This is problematic and unrealistic for several reasons. First, Huber beds are seldom rented. The Sauk and Columbia County seats are close enough that the two counties might have been able to collaborate on this, but both counties built new Huber facilities instead. Secondly, industry standards consider a jail to be full when 80% of the beds are occupied. That point was made by the state jail inspector at the Feb. 8, 2005 "jail meeting." The sheriff’s department has made this point repeatedly. We shouldn’t count cells we don’t intend to use for long term use to make the per unit cost look lower than it will be. Between the two factors, it should be a long time before the average daily population (ADP) exceeds 300 in the two units combined.

   Thirdly, knowing the daily cost per inmate day of owning and operating a jail is very useful in deciding a fair rent rate to charge, or in making the decision to rent versus build. Pretend for a moment that we are far enough into the future that Sauk County has one more inmate than needed to "fill" the Huber unit and unit B. This could cause some to favor opening the empty unit. It would also raise the question about the cost of incarceration alternatives. Further pretend that by this time, other countries have built jails and instead of wanting to rent from other counties they are willing to rent to other counties such that it would cost Sauk County $58 an inmate day including transportation. Finally, pretend that the added first year cost of operating half of the empty unit is the $614,175 added cost currently projected (by proponents) to open half of the empty unit. In these circumstances, the daily cost of using the empty unit would be $1,683 = ($614,175/365) enough to rent space for 29 inmates at a daily rate of $58. It would be terribly misleading to divide the daily cost of $1,683 by the 47 beds and conclude that the daily cost per inmate is $35.80. Clearly, with one inmate in the unit and a total daily cost of $1,683, the daily cost per inmate would be $1683. Even with the unit half full, it is more informative to calculate the cost per inmate instead of the cost per bed. Yet with the Huber unit and unit B only about half full of our inmates, their analysis calculated cost per bed and didn't calculate cost per inmate. This should illustrate why it is so dangerous to calculate the cost per bed of unit B and the Huber unit by dividing the daily net cost by the number of beds without also dividing the daily net cost by the number of Sauk County inmates (the 2004 average daily population of Sauk County inmates was146)..
    

   Rent Rate Ceiling and Floor
   Now let’s pretend that Dane County was using this analysis to help decide whether or not to build a 369 bed jail to house 146 of their inmates. As part of the decision, the county should compare the cost of incarcerating in owned versus leased space. It is likely that if Dane county decided to build a 369 bed jail now, it would cost more to build and operate than the cost of the Sauk County jail, mainly because wage rates in Madison are likely higher than in Baraboo. Using an average rent rate of $52.00 per inmate day plus $6.00 transportation cost per inmate day it would cost a total of $3,090,820 (146 inmates times $58.00 rent times 365 days) to rent space for compared to the total annual net cost of $5,347,516 or gross cost of $6,693,216 calculated in the analysis (a range of $100 to $125/inmate day). Compared in this way, renting is much more economically attractive than building.
   
   If Dane County thought the $39.70 net cost per bed per day from the analysis meant the cost per inmate day, they could erroneously conclude that building a 369 bed jail was more economical than renting space for 146 inmates at an inmate rent and transportation cost of $58 a day. This would be a serious mistake.

   The behavior of Dane County’s policy makers indicates that they understand this better than the policy makers in Sauk and many other counties.

   Until challenged, proponents of opening the empty unit recently claimed that Dane County had no where else to go with their inmates. If proponents were correct in saying Dane County had no where else to go with their inmates, then building their own facility is their next best alternative to renting from Sauk County. Dane County likely knows that it would cost them from $100 to $125 per inmate day in a new facility they build. A bit less than $100 to $125 per inmate day would be Dane County’s rent price ceiling. The rent rate floor ought to be more than the added cost per inmate day. The actual rate will then be determined by supply and demand and the knowledge and negotiating ability of the "renter versus the owner."

   I argue that a more fair rent level between 2 parties, each wanting to be relatively fair and each knowing the level of the floor and ceiling is about half way in between. $50 is close to the floor. Half way between a floor and ceiling of $50 and $120 is $85.

    

   This should make it clear that calculating the gross or net cost per bed may be interesting, but as the only or primary measure, it is far less useful than calculating the daily cost per inmate. The analysis would be far more useful if it included a daily cost per Sauk County inmate calculation for all cost factors along with other measures as I did in my analyses. Doing the calculations at various levels of occupancy would be useful too. Again, I did this in my analysis. If we divide the $5,347,516 net cost calculated for 2005 by the 2004 ADP of 146, we calculate an average cost of about $100 per inmate day.
   
   Some people will want to divide the total daily net cost by the combination of Sauk County inmates and rental inmates. This would only be correct if the renting inmate’s rent was always adjusted to equal the cost to Sauk County per inmate day for housing Sauk County inmates. However, in the act of subtracting all rental income from gross cost, the impact of rental inmates is already accounted for. Using them in the divisor clearly double counts their impact and results in an erroneously low estimate of cost per Sauk County inmate day.

   It was clear to me that at least one member of the finance committee (at our brief meeting on March 15) was thinking that the cost per bed was the cost per inmate. If one thought the net cost per bed per day from the analysis meant the cost per inmate day, one could make some serious mistakes.

2. A second problem with the analysis, as it is presented, is that it assumes an 80 year depreciation period. If it is appropriate to use an 80 year period for the current jail, then Sauk County has 35 years of depreciation left on the old jail (which was just dismantled) and about 60 years left on the original Huber unit.
    
3. This third problem is related to point two. The 2005 Sauk County budget book on page 376 (and page 134 of the proposed budget) documents that the cost of debt service on the law-enforcement center will be $1,761,100 in 2005. It also documents that utilities and routine maintenance is budgeted to be $702,186 in 2005. Together these two items equal $2,463,286. Yet the analysis uses $299,071 for utilities and maintenance, $144,812 depreciation and $446,597 for a total of $890,480 or about 36% of $2,463,286. If the diagram of the Sauk County Law Enforcement Center presented in the packet with the analysis is drawn to scale, then it is clear that the parts of the building that are for jail functions or mainly for jail functions (such as kitchen) represent far more than 36% of the complex. In fact, an eyeball analysis suggests that the jail represents at least 80% including the empty unit, but excluding the Sheriff Department offices, the Coroner’s Office and garage as not jail). Especially when one considers that office space costs much less to build than it costs to build cells with all the command center technology, I assumed jail functions represented 80% of the cost of constructing and operating the law enforcement center when I did my budgets. My budgets include 80% of the total debt service, utilities, maintenance. Using depreciation in such a case would only duplicate the principal part of the debt service. If the jail is charged with 80% of these costs, the cost per inmate day would increase by $20.82 and cost per bed by $8.24. It also increases the gross and net cost by $1,110,000 for a total of $7,803,216 annual gross cost and $6,457,516 annual net cost.
4. The cost per bed analysis is based on the employee compensation levels of 2005. Employee compensation rates have increased by about 5% compounded annually. I also used the then current wage rates when I did my cost per inmate analysis. At that time, I also didn’t factor in the employee compensation rates of later years because not only the cost per inmate day but also the cost per bed (at 80% capacity) was very high – much higher than the market rental rate. My budgets did show how much higher employee compensation could be in five and ten years. Over a 20 year period, employee compensation can easily represent 70-80% or more of the cost of owning and operating a jail. Under such a cost structure, it seldom pays to construct substantial excess capacity very far in advance. No smart farmer builds a barn now for cows he may have 10 years from now.
5. Not only does the analysis understate the jail’s share of debt service utilities and routine maintenance, it doesn’t provide for the cost of more major repairs. There shouldn’t be major repairs in the first few years. I’ll guarantee there will be major repairs in the first 20 years and in years after that. I used 3% of new cost as an annual estimate. This actually is a fairly small item compared to the previous points.

1. Increasing the inmate population in the staffed parts of the jail will usually decrease the cost per inmate but increase the total cost and the cost per cell.
2. At any percent capacity, increasing the proportion of renting inmates decreases the cost per inmate and net cost assuming rent rates are high enough.

1. How many staff people, jailors, booking, classifiers, nurses, and administrators are employed in each jail unit?
   - Huber – female
   - Huber – male
   - Unit B

With a ratio of 8 to 10 males for every female in Wisconsin jails, why is half of the capacity of the Huber unit designated for females?