Another Cost of Sprawl
The Effects of Land Use on Wastewater Utility Costs
GEOGRAPHICALLY SENSITIVE USER RATES: THE CLEVELAND EXPERIENCE
Cleveland's water supply system utilizes a unique rate structure by which customers in its higher elevation areas (which tend to be located the furthest from the urban core) pay more on a per-unit basis for water than do customers in its lower elevation areas. As discussed below, these rate differentials are apparently intended to recovery additional pumping costs associated with higher elevation (and to a lesser extent, distance) rather than the sorts of differences associated with service densities, as suggested in Part 1 of this study. Because outer suburbs are typically less dense than inner cities, however, CDOW's rate structure begs the question of whether it is an appropriate surrogate for relieving the cross-subsidies associated with systems that serve different densities.
A thorough examination of this question is rendered exceptionally difficult, unfortunately, by a lack of financial and system data segregated by community or neighborhood within CDOW's service area. (This is not unusual among large regional water service providers.) Moreover, an attempt to construct an accounting from existing data that would permit such an examination is, if possible at all, well beyond the resources available to this study. For these reasons, readers should use caution in interpreting the findings in Part 2.
Nevertheless, a rough analysis comparing in-city and out-of-city costs, along with in-city and out-of-city rate differentials, was performed to give an idea of whether rates generally can be said to reflect elevation and distance-related differences between the costs of providing water to city customers and those of providing water to suburban customers. In general, it can be said that they do. In this sense, other water providers might do well to emulate Cleveland's differentiated rate structure in order to more closely recover additional costs associated with service to outlying areas.
What is also apparent, however, is that the rates do not attempt to recovery - nor do they inadvertently recover - the differences in costs attributable to different land use patterns, including such differences that existing within zones where customers pay the same rate for water. Because these differences can be substantial, as revealed in Part 1 of this study, it is likely that customers in denser neighborhoods with CDOW's service area are subsidizing customers in neighborhoods that are less dense.
The Cleveland Division of Water (CDOW) provides water supply, treatment, storage and distribution for virtually all of Cuyahoga County, Ohio (including Cleveland) and parts of Geauga, Medina, and Lake Counties. In addition to Cleveland, CDOW provided service to 64 "direct service" suburbs, six master meter suburbs, and three master meter counties in 1996.  CDOW provides maintenance service on all major waterworks facilities, including the repair and maintenance of all trunk mains and most distribution mains in the direct service suburbs.  It also performs customer billing for most of the direct service suburbs. Master meter suburbs and counties purchase treated water from CDOW and are responsible for the maintenance and operation of local distribution lines, service connections and customer billing within their respective service areas.
CDOW is one of the few utility systems in the U.S. that charges rates based on geographic service cost differentials. CDOW's service area consists of four pressure zones corresponding to elevation.  Zones with higher elevation require additional pumping, pressurization, and related capital facilities. These facilities in turn cause CDOW to incur additional O&M costs. CDOW's existing rates recognize that distribution costs for these areas are higher due to their higher elevation.
Similarly, the rates recognize that areas with lower elevation, and thus lower pressure requirements, are less costly to serve. The lower pressure zones are also generally closer to CDOW's treatment plants. This would also cause them to be less costly to serve. In CDOW's case, elevation and distance are closely related since the higher elevation areas also tend to be furthest from its treatment plants. It is the combination of these two factors which were presumably considered when the existing rate differentials were adopted in the late 1970's. There is no indication that CDOW's zone-based rates reflect cost differentials attributable to economies of scale or density.
Research cited in Part 1 indicates that size (for treatment) and density (for distribution) are two additional factors which can influence utility service costs. Customers in the City of Cleveland account for the largest portion of CDOW water usage by far, making them primarily responsible for providing favorable economies of scale for CDOW's treatment functions.  And, since the City along with other of its suburbs are the highest density areas within CDOW's service area, they would also tend to be the least costly to serve insofar as distribution costs.
Most of the recent growth within the Cleveland metropolitan area has occurred in CDOW's two highest pressure zones, the second and third high zones. Not only do communities located in these zones require additional pumping facilities and major transmission lines due to elevation and distance, some of them are among the least dense in the region. Recent development activity in CDOW's two higher pressure zones has also required CDOW to construct additional treatment capacity at the Crown treatment plant, even as water usage by customers in the City some inner-suburbs has declined.
CDOW'S RATE STRUCTURE
CDOW's current rate structure was established in the 1970's as a negotiated resolution to a lawsuit between the City and several suburban communities. The dispute involved how the region's water service was to be provided and paid for. The suburbs needed a reliable source of water, but the City owned the water system. Due to its financial difficulties, the City was unable to simultaneously finance needed improvements required to serve the suburbs, and make repairs and upgrades required to keep the existing system operating properly. The resolution was for the City to maintain ownership of the system, with the suburbs agreeing to the zone-based rate structure. In exchange, CDOW was to continue to provide water service to all suburban jurisdictions by agreement between each suburban community and the CDOW. In return, the suburban communities agreed not to challenge CDOW's rates for a period of ten years.  (Appendix B contains pertinent language from Articles 4 and 5 of CDOW's standard direct service suburb water service agreement which incorporate this arrangement.)
Rates for customers located in CDOW's direct service suburbs are, by agreement, charged as a multiple of inside-city rates, and all inside-city customers are charged the inside rate regardless of the pressure zone in which they are located. The rate multiples are as follows: 1.75 for the low and first high service districts outside the City of Cleveland; 2.00 for the second high district; and, 2.30 for the third high district. As a result of these rate differentials, in 1996, CDOW customers located within the City of Cleveland accounted for 40.3 percent of metered water consumption, yet generated only 26.3 percent of metered water charges.
The effect of these rate differentials can be seen by comparing the effective inside-city and outside-city user rates (the effective rate is metered sales revenue divided by metered consumption). This serves as a basis to compare the relative source of user charges between the City of Cleveland and the balance of CDOW's service area. In 1996, the effective rates were $10.67 per thousand cubic feet (MCF) for inside-City customers, and $21.17 per thousand for customers located in CDOW's direct service suburbs, for a 1.98-to-1 outside to inside ratio. The overall outside-to-inside rate differential appears to roughly correspond to cost differentials attributable to higher elevation-related costs for CDOW's higher pressure zones. (See Appendix E for a detailed explanation of how this was verified.)
For example, the majority of CDOW's capital improvements over the 1990 to 2001 time period are to provide service in the outer suburban areas of Cuyahoga County, and parts of neighboring Geauga and Medina Counties. During this period, CDOW's capital improvements will total about $725 million, of which about 65 percent are estimated to provide service outside the City of Cleveland and 35 percent are to provide service within the City of Cleveland.  This is about a 1.85-to-1 ratio between outside- and inside-city improvements and is very close to the existing effective overall rate differential.
And, when average cost techniques are used to allocate CDOW's O&M costs based on consumption, pumping capacity, miles of pipe and employment, an outside/inside-city rate differential of about 1.85-to-1.00 also appears to be justified. Assuming the difference between the effective 1.98-to-1 inside/outside rate differential and the 1.85-to-1 ratio (as estimated in Appendix E) is attributable to a combination of estimation error and/or other variables, it would appear that, at least on a large scale, the average cost differences between serving in-city and out-of-city customers are generally reflected in the rate differentials. It does not necessarily follow from this observation, however, that either density or economies of scale are currently reflected in CDOW's rate structure.
SERVICE DENSITY AND ECONOMIES OF SCALE WITHIN CDOW'S SERVICE AREA
There are substantial variations in service and grid density within CDOW's service areas. For example, about 1,350 miles of pipe are estimated to be required to provide service within the City of Cleveland. There are currently about 680,769 service units in the City of Cleveland.  This yields an average inside-City service density of about 504 service units per mile of pipe.  About 3,650 miles of pipe provide service to CDOW's direct service suburbs. These suburbs have a combined total of about 989,000 service units, yielding an average service density of about 271 service units per mile of pipe. 
Service density differentials are even greater for individual direct service suburbs. An alternate method of estimating service units per mile of pipe results in a slightly higher service density for the City of Cleveland, but allows a consistent comparison to other Cuyahoga County communities to illustrate a range of differing service densities. In particular, Table 2 provides an approximation of the range in service density for communities served by CDOW in Cuyahoga County using centerline miles (CLMI) as a proxy for miles of pipe for communities for which this data is unavailable.  By this measure the average service density for direct service communities in Cuyahoga County (inclusive of Cleveland) is about 507 service units per estimated mile of pipe.  Approximate service densities for individual communities range from a high of about 791 units per estimated pipe mile in Brooklyn to a low of about 70 in Cuyahoga Heights, with Cleveland at 546. The thirteen most dense communities, with a total of 1,116,988 service units, have an average service density of about 607 service units per estimated mile of pipe.  The thirteen least dense communities, with a total of 24,725 service units, have an average service density of approximately 120 service units per estimated mile of pipe. 
CDOW's service area in Cuyahoga County, excluding Cleveland, has an estimated service density of about 414 service units per mile of pipe, making the City about 32 percent more dense than the balance of CDOW's direct service suburbs in Cuyahoga County. 
Based on the results of Part 1 of this report, the wide range of service densities within CDOW's service area would suggest that on the basis of service density alone, unit distribution costs are likely to be higher in some parts of CDOW's service areas, and lower in others. Since it appears that CDOW's current rate differentials do not consider density-related service cost differentials, customers in CDOW's higher density communities are likely subsidizing distribution costs for customers in its lower density communities.
In CDOW's case, its higher density communities (particularly Cleveland) also have higher poverty rates and lower per capita incomes than its lower density communities. In the thirteen highest density communities (the top quartile), an average of 19 percent of all persons were below poverty status. For the thirteen lowest density communities (the low quartile), only 3.3 percent of all persons were below poverty status. Similarly, per capita income in the high density quartile averages $18,209, and $30,876 for the low density quartile. Subsides from higher density/lower income communities to lower density/higher income communities would exacerbate disparities between these communities.
|Table 2: Approximate Local Service Density for Cuyahoga County Direct Service Communities|
|Community (Rank by Density, 1=Highest)||Population||Employment||CLMI or PM||SU's Per PM|
|Bay Village (33)||17,000||980||62.3||288.7|
|Bedford Heights (13)||12,131||9,079||40.7||521.1|
|Broadview Heights (30)||12,219||2,927||46.4||326.5|
|Brook Park (2)||22,865||18,070||56.0||730.9|
|Brooklyn Heights (41)||1,450||NA||10.8||134.9|
|Cuyahoga Heights (50)||682||NA||9.8||69.9|
|Fairview Park (24)||18,028||3,197||53.8||394.5|
|Garfield Heights (20)||31,739||7,039||91.4||424.1|
|Gates Mills (49)||2,508||144||31.5||84.3|
|Highland Heights (25)||1,934||4,829||17.2||392.3|
|Highland Hills (26)||1,934||NA||5.0||386.8|
|Hunting Valley, est. Pop (46)||1,458||NA||15.4||95.0|
|Maple Heights (14)||27,089||7,416||69.0||500.0|
|Mayfield Village (17)||3,462||3,092||13.7||478.0|
|Mayfield Heights (3)||19,847||8,891||39.5||726.8|
|Middleburg Heights (4)||14,702||10,520||39.7||635.2|
|Moreland Hills (43)||3,354||205||29.4||120.9|
|Newburgh Heights (31)||2,310||NA||7.3||316.4|
|North Olmstead (35)||34,204||9,797||175.4||250.9|
|North Royalton (27)||17,671||4,114||59.2||367.9|
|Oakwood Village (36)||3,392||1,805||22.5||230.6|
|Olmstead Falls (19)||6,741||792||17.2||438.2|
|Orange Village (37)||2,810||511||14.6||227.2|
|Parma Heights (7)||21,448||3,688||44.1||569.7|
|Pepper Pike (39)||6,185||807||39.4||177.5|
|Richmond Heights (6)||9,611||3,766||21.9||611.1|
|Rocky River (32)||20,410||6,453||90.0||298.5|
|Seven Hills (21)||12,339||1,389||32.6||421.2|
|Shaker Heights (28)||30,831||4,517||97.1||364.0|
|South Euclid (18)||23,866||4,125||61.5||455.4|
|University Heights (12)||14,790||1,841||31.5||528.0|
|Valley View (44)||2,137||NA||21.3||100.4|
|Walton Hills (45)||2,317||NA||23.7||97.9|
|Warrensville Heights (16)||15,745||4,013||40.9||483.3|
|Average excluding Cleveland||414.5|
Similar differences exist when one looks at economies of scale. Customers located within the City of Cleveland account for about 40 percent of CDOW's water usage. The next single largest source of water usage is the City of Parma, which accounts for only 5.67 percent of CDOW's water usage. This indicates that customers located within the City are primarily responsible for favorable economies of scale for CDOW's treatment plants. If CDOW's rates do not recognize the favorable economies of scale provided by the City of Cleveland, then customers located inside the City are likely to further subsidize those located outside the City. 
Similar to any subsidies resulting from service density, the direction of subsidies resulting from economies of scale is from lower income, higher density communities to higher income, lower density communities. For example, Median Household Income (MHI) in Cleveland is $17,822, compared to $28,595 for all of Cuyahoga County (inclusive of Cleveland). And, the number of persons in poverty status in Cleveland is 28 percent, while for the balance of the County it is only 5.4 percent. 
CDOW'S WATER SERVICE AGREEMENTS
As demonstrated above, service density and economies of scale likely play a role in service costs within CDOW's service area. One barrier towards incorporating density and scale economy variables into CDOW's rate structure is the existing water service agreements. Originally drafted in the late 1970's to implement a negotiated settlement involving a regional dispute over the provision of water service, these agreements provide that CDOW's direct service suburbs will not challenge its rates so long as the rate differentials agreed to in the late 1970's are maintained.
Shifts in CDOW's service area population, employment and land use characteristics have, and will continue to affect its costs. While CDOW's existing rate structure recognizes that its higher elevation areas are more costly to serve, they fail to recognize the effects of economies of scale or density as a factor. The foregoing analysis suggests that incorporation of a density and/or an economy of scale variable within CDOW's rates should be considered based on cost-of-service principles.
To do so would require re-opening the issue of how much CDOW's direct service and master meter suburbs should pay for treated water and for water distribution services. This would require re-negotiation of CDOW's water service agreements with each of its direct service suburbs prior to their expiration. Prior to pursuing this course of action, it would be advisable to prepare a detailed cost of service analysis using actual service records and other data not available for this study to more precisely understand the nature of water service cost differentials within CDOW's service area.
30. Three Cuyahoga communities operate their own distribution systems.
31. Geauga, Medina and Lake are "master meter" communities. Master-meter communities are responsible for operating their own distribution systems. "Direct service" communities are those in which CDOW is responsible for both providing water and for distribution system operations. Medina and Lake County service is for emergency stand-by service only.
32. Official Statement of the City of Cleveland, Ohio Waterworks Improvement and Refund Revenue Bonds, Series H, 1996, May 1, 1996, p.26.
33. These are referred to as the low, first, second and third High zones.
34. The same would likely hold true for source of supply functions the costs of which also are favorably influenced by economies of scale (size).
35. Sample Direct Service Suburb Water Service Agreement provided by CDOW, Articles 4 - Water Rates, and 5 - Covenant Not to Sue on Water Rates; Exclusive Franchise.
36. Data available for this study allowed only a broad analysis of the overall inside/outside cost and rate differentials. To confirm this finding, and to determine whether cost and rate differentials are in balance for each community within CDOW's four pressure zones would require a detailed inventory of the system's distribution lines, engineering criteria and cost review.
37. Review of CIP by Mr. Chris Koscan, CDOW CIP manager, 1997. These improvements, however, may not have the same impact on CDOW's O&M expenses. Most of the inside-city capital improvements are to upgrade the existing conveyance system. As such, they are likely to lower costs since they will result in the replacement or repair of some of the oldest and most costly portions of CDOW's conveyance system. This is not the case for many of the improvements planned to provide service outside the City, many of which are entirely new facilities or substantial expansions of existing facilities (such as the Crown treatment plant expansion) that will require net new O&M outlays.
38. 1994 estimated population of 492,901 plus a total of 187,868 manufacturing (1995 data) and wholesale, retail and service jobs (1992 data).
39. This includes pipe for local service and the estimated share of major distribution lines and interconnects for inside-City service. When considering only pipe for local service, the City's estimated service density is about 546 service units per mile of pipe. However, on an equivalent basis, the number of service units are likely to be higher in the City of Cleveland with its higher concentration of water-process intensive industry.
40. The miles of pipe assigned to the direct service suburbs may not include recently installed pipe.
41. Ohio Department of Transportation (ODOT), Municipal Street Mileages as of January 1, 1997, p. 18. Miles of pipe for Cleveland and each of CDOW's direct service suburbs was unavailable for this study. CLMI in each community is used as a proxy for miles of pipe. Actual service density for communities with significant new development in the last ten years will be lower than that shown since ODOT's CLMI data has not been updated in ten years.
42. Includes major distribution lines, and interconnects. Local service density, which includes lines required for local service only is greater based on CLMI data.
43. Brooklyn, Brook Park, Mayfield Heights, Middleburg Heights, Beachwood, Richmond Heights, Parma Heights, Solon, Cleveland, Strongsville, Parma, University Heights and Bedford Heights.
44. Woodmere, Pepper Pike, Bratenahl, Brooklyn Heights, Linndale, Moreland Hills, Valley View, Walton Hills, Hunting Valley, Bentleyville, Glenwillow, Gates Mills and Cuyahoga Heights.
45. Due to loss of employment and population, current service density within the City is substantially less than in the past. Between 1980 and 1990, population within the City declined by about 12 percent from 573,822 persons to 505,616. Employment has also declined substantially within the City. For example between 1980 and 1990 the City lost 3,285 non-manufacturing jobs, and between 1990 and 1995 it lost 11,800 manufacturing jobs.
46. Including Cleveland. Based on 1990 population and manufacturing employment, and 1992 wholesale, retail and service employment. Miles of pipe for each community was unavailable. Miles of wastewater pipe was used for the six communities for which it was available as it ordinarily corresponds closely to miles of water pipe. CLMI data is used for other communities in lieu of pipe data. Average service densities in the non-Cuyahoga portions of CDOW's service area are likely below those for Cuyahoga County.
47. SU's equals population plus employment. PM is estimated pipe miles using either CLMI or wastewater pipe miles. This measure may understate equivalent density for communities with manufacturing employment as these facilities tend to be more water intensive.
48. Customers located inside the City bear the risk of ownership of the system. Rate-design principles recognize this as an additional basis for charging non-owners (outside-city customers) higher rates to compensate for the risk premium borne by inside-city customers. The study does not consider a risk premium which, if included, would justify an even greater outside/inside rate differential.
49. 1989 data. The MHI data for Cuyahoga County includes Cleveland, and is significantly lower than it would be if Cleveland were excluded.
Sign up for NRDC's online newsletter
Kaid Benfield's Blog
Kaid Benfield writes about development, community and the environment on Switchboard.
Kaid's Recent Posts
- New guide from ASLA highlights urban green, literally, in Portland
- posted by Kaid Benfield, 10/27/14
- Compact and connected communities improve public health, says new research
- posted by Kaid Benfield, 10/20/14
- For smart growth, not all urban density is created equal
- posted by Kaid Benfield, 10/6/14
NRDC Gets Top Ratings from the Charity Watchdogs
- Charity Navigator awards NRDC its 4-star top rating.
- Worth magazine named NRDC one of America's 100 best charities.
- NRDC meets the highest standards of the Wise Giving Alliance of the Better Business Bureau.