Budgeting Strategically

Since the profligate 1960s, most institutions of higher education have struggled to fulfill their academic mission in the face of increasing costs and increasingly constrained resources. Many have turned away from traditional budgeting to models focused on aligning incentives more strategically.

Under the traditional budget process, especially in the public sector, the emphasis lies on a highly centralized annual exercise that looks at a relatively fixed base as a given and adds an inflationary component across the board with an emphasis on “fair shares” for everyone. Harvard served as an outlier for many years for what became known as “Every Tub on Its Own Bottom.” This meant a decentralized structure in which every academic unit, such as a college, functions as a somewhat independent business responsible for both its own revenues and costs. As financial pressures mounted, other institutions developed their own versions of this approach. The University of Pennsylvania (Penn) adopted what it called Responsibility-Centered Management (RCM) in 1974. The University of Indiana at Indianapolis and Bloomington was the first public university to convert to RCM in 1990. ^[1]

Since then, alternatives to traditional budget models have spread, some under the RCM banner and others under a variety of titles, including Value-Centered Management, Budget Restructuring, Budget Reform, The Budget Model, Performance-Based Budget, Activity Based-Budgets, and Hybrid Budgets. These alternative models now account for a third to a half of all public and private institutions, including many of the top universities in the country. ^[2]

Although these all have some RCM-type characteristics in common, finding a label that fits a variety of institutions and approaches can be tricky. For example, sometime in the late 1990s I was part of an Ohio State University delegation that visited the University of Indiana at Bloomington to learn more about their RCM system. We asked them what they thought was the single biggest improvement from the old traditional system. They agreed it was the ability for academic units to carry over unspent cash balances at the end of the year. We were surprised, because Ohio State’s system, which was very traditional at the time, had permitted the carry-over of cash balances for years.

Rather than devote a lot of time on definitional issues, in this essay I will approach what I call “Strategic Budgeting” by identifying best practices that apply to both public and private institutions. I call this approach “Budgeting Strategically” and conduct my evaluation around these three building blocks:

1. Aligning budget with academic goals.
2. Ensuring financial sustainability.
3. Achieving campus wide buy-in.

Aligning Budget and Planning

Critics of traditional budgeting contend that it places too much emphasis on an annual accounting exercise and not enough on a multiyear strategic advancement exercise.^[3] If you go on the website of almost any college or university, you will likely find a strategic plan, often in great detail. You will also find some sort of budget document that lists sources and uses by accounting code, often in great detail. Very seldom will you find any document that links the two. But a budget without a long-term plan is like a car without a driver. A plan without a budget is like a car with no fuel.

Declaring a priority is one thing. Funding it is another. Critics of traditional budgeting argued that the bottom up, historic fair shares characteristic of most traditional systems squeezed out resources for needed investments, and that by the time historic budget bases received an inflationary increment and their fair share under traditional budgeting, needed investment directed at the future was left on the cutting-room floor.

Supporters of a more strategic approach argue this need not be a huge amount of money, maybe 1–2 percent of annual spending, but it should be reserved for high priority strategic initiatives that fundamentally improved the organization’s future prospects. This meant a wide number of units would be eligible, but awards would be competitive and not allocated as “fair shares” (Jones 1995, 11-12).

Those colleges and universities who have moved away from traditional budgets approached this differently, but most have set aside central resources for future investments. This includes the early RCM adopters like Penn, Indiana, and Michigan, but also later adopters of different systems such as the University of Arizona, UNLV, and Kent State University. A 2015 review by the Education Advisory Board concluded that colleges and universities with non-traditional budget systems allocated a median budget of 1–3 percent of annual revenues for central funding of strategic initiatives (Education Advisory Board, slide 50).

That said, very few institutions have specifically aligned the strategic plan priorities and budget-funding priorities in their budget documents. Ohio State University is one that currently does in the form of a four-page narrative as part of its formal budget presentation to the board of trustees (Ohio State University Financial Planning and Analysis 2023).

When Ohio State converted in the early 2000s, it specifically set aside funds for strategic priorities. Five percent of the annual increase in tuition and state support was reserved for centrally budgeted strategic initiatives, including competitive awards to academic departments under the Selective Investment initiative.

A decade later, the university built on this experience to launch the Discovery Themes initiative, which was both more expansive and more focused than what had happened before. It was more expansive in that any unit could apply, including centers. It was more focused in that it was limited to three thematic areas where the university had already established a strong reputation in teaching and research. These were Health and Wellness, Food Protection and Security, and Energy and Environment. A fourth related to the Arts and Humanities was added later.

Another significant change was to encourage proposals that crossed disciplinary lines where often the most exciting and significant research was likely to take place. This included areas like biomedical engineering, infectious diseases, and alternative energy sources. The response across campus was extremely favorable, and the university received many more worthy proposals, particularly of an interdisciplinary nature than it could support (Asher 2021).

The university’s leadership also concluded that the 5 percent set aside established as part of the 2003 budget restructuring was insufficient to produce the results expected. Even with annual General Fund revenues growing at 5 to 6 percent, that meant 5 percent of that was only 0.025 to 0.030 percent of the total annual spending, well below the 1 to 2 percent desired. So in 2012 the university decided to monetize its parking assets by leasing them out to a private company. This move created some controversy on campus over the concerns about lack of control over an important asset, but in exchange the university received a lump-sum payment of $483 million. Instead of spending it outright, the bulk of the money was invested in the university endowment where it could be expected to pay out more than $20 million annually in perpetuity and increase faster than inflation (Asher 2021)

Up to this point, we’ve addressed the importance of the 1 to 2 percent of spending that supports key initiatives, but what about the remaining 98 to 99 percent?

Incentives

Although the continued funding of campus-wide strategic initiatives is critical to the future of any college or university, that is not a reason to ignore the more routine day-to-day operations. Traditional budgeting in a highly centralized system based on historic shares provides very little in the way of financial rewards for deans and other academic leaders. They may not be part of central administration but should be treated as an important resource to develop new or expanded programs in response to a changing environment.

This became very clear to me when a senior university official responded to a question I posed about our stubborn closed-course problems shortly after I arrived at the university in 1990. He told me they did not take enrollment shifts into account in making budget allocations because enrollment fluctuations were based on fickle student interests. He used as an example nuclear engineering, which was popular in the 1950s and 60s, but fell out of favor after Three Mile Island in 1979.

Three years later, I was at a meeting with the deans where we talked about how to regain our budget footing after a series of state-imposed budget cuts. Our provost asked the deans why they weren’t more aggressive in developing new or expanded programs to attract more students. They responded they lacked the confidence that central administration would reward them in the budget process for the risk they took in committing resources to something new or different.

Ohio State’s challenges at this juncture might have been more extreme, but not unique among centrally controlled budget systems. Thus the idea of a decentralized system with a set of financial incentives empowering academic leaders at the college level to have a greater say in resource allocation as responsibility centers began to take hold. Under this system academic units (usually colleges or schools) would be guaranteed a significant portion of the income they generated, but would also accept a significant portion of the costs they incurred.

Revenues

Student tuition is the largest source of support for instruction for most colleges and universities. In higher education costs, such as salaries and utilities, tend to be fixed at least in the short run; therefore, the ability to sustain or grow enrollments is essential to advancing the institutional mission.

Under the more traditional budget, all tuition and state unrestricted support is collected centrally, then distributed to academic and academic support units, usually based on some sort of historical share. This has the advantage of stabilizing funding, but at the expense of incentives to adapt or innovate. We have already discussed the problems created, such as closed courses and lack of incentives to develop new programs, when tuition doesn’t follow students.

Once a college or university decides to distribute tuition revenue to its academic units in a non-traditional way, it needs to decide four things: who those units are, how to measure enrollment, what time frame to use, and how to fund support services, such as libraries, which are important but don’t directly generate revenue. Various institutions approach this differently, but best practices cluster around using colleges or schools as the responsibility units, with a two- or three-year moving average of enrollment as a base. Credit hours are the most commonly used measure of activity, but this has broadened in recent years as the public has called for greater accountability for timely graduation and course completion. Many institutions now use a combination of credit hours, majors. and degrees granted (Education Advisory Board, slides 30-34). An RCM-type formula also provides a transparent way for colleges and universities to distribute the increased revenue from new programs such as the increased number of online offerings.

Expenses

Wages and salaries are the single largest expense for academic units. Under both traditional and incentive systems those expenses are charged against a unit’s budget. In a traditional budget system, units are expected to keep their expenditures within the amount determined centrally and heavily influenced by history. Under the incentive-oriented systems, units are also expected to manage within their budgets, but those budgets are determined more by revenues generated by the unit rather than a central determination.

In a traditional budget other functions, such as space, student support, and libraries, are funded centrally. Under the more incentive-based system, budget units are also expected to pay all other expenses associated with supporting their units. This can include direct charges for things like employee benefits or space, specialized support services like student financial aid, the registrar, library, and public safety. Every college and university does it a little differently, but most use a mix of direct charges, charging algorithms, and general taxes or assessments.

Budgets for these items are determined centrally, but costs are often allocated back to the responsibility centers. Space and employee benefits are often charged directly. Algorithms based on the number of students taught are usually used to fund student support services such a financial aid, scheduling, and so on. The remainder, which might include the library, public safety, and the president’s office, are funded through a general tax or assessment on revenues, usually at a rate of between 20 and 30 percent (Education Advisory Board, slides 35-40).

Reconciliation

Once the formulas for distribution of revenue and expense are determined, the next step is to run them though with live numbers to see what the bottom line looks like for the individual colleges and schools. Most universities transition to a new system by initially producing a dry run for a year as a learning exercise to allow everyone to become familiar with the new budget system before moving ahead to full implementation.

Officials at Ohio State took this one step farther in a unique effort to explicitly reconcile both the budget numbers and the academic goals they are supposed to support. The provost did this via a twenty-five-page memo directed to the deans but shared with everyone on campus. He began by laying out the six priorities of the university’s academic plan and how the new budget system would support them. He then went a step farther by explicitly listing the thirteen selective Excellence Programs recognized by the university through a rigorous peer review process.

The university administration had been reluctant to single out these programs in the past, fearing it would paint the other programs as losers. But at this point the leadership concluded that if the university were to live up to its aspirations, it would have to learn how to celebrate excellence without demeaning other units.

The provost laid out the projected base budget numbers for each of the eighteen colleges in the proposed new budget system in the second part of the memo. These were labeled “Sources and Uses Statements” as opposed to profit and losses to deliberately stay away from business jargon. These statements showed that eight colleges generated more revenue than expense to support the other ten who did not.

In the third part of the memo the provost reconciled the academic goals described in part 1 with the budget realities in part 2. He emphasized that he was not moving to an “every tub on its own bottom” system. Instead, he divided the eighteen colleges into three categories and described how they would have their base budgets held harmless, decreased, or increased over the next five years.

This was a somewhat risky step politically because it was so explicit, but in fact it received a great deal of support from the deans and other stakeholders, including the faculty, because it provided a sense of clarity and reason to what had previously been viewed as an opaque process. ^[4]

Sustainability

One of the principal weaknesses of traditional budgeting is the tendency to focus on only one year at a time. With relatively high fixed costs in the short run and revenue sources such as tuition and state support that may fluctuate without much warning, higher education institutions need to be prepared to sustain themselves in both good times and bad.

Reserves

Financial sustainability begins with some sort of reserve or rainy day fund to help buffer against unpleasant surprises. Most public and private colleges and universities usually end the fiscal year with a relatively large amount of cash on hand, but unless structured properly it can produce a false sense of security. This is one-time money, so once it’s spent it’s gone. And as it accumulates, it takes on the appearance of a slush fund that creates political problems for both internal and external constituencies.

The best way to counter this is to develop a clear, coherent set of policies that lay out what should be set aside and how reserves should be used. These tend to break out into one of three categories:

1. Operating/Contingency. These are the funds to ensure that bills get paid on time and to offset an unexpected drop in enrollments, loss of state support, or other unpleasant surprise. The National Association of University Budget Officers recommends that this amount should equal at least 25 percent of annual expenditures ^[5]
2. Infrastructure. This is money set aside to renew critical infrastructure, including aging facilities and in many cases technology as well. These amounts vary greatly by institution.
3. Strategic Initiatives. I discussed this earlier in the section titled Aligning Budget and Planning. This is a relatively small amount (1 to 2 percent of annual spending) but needs to be funded consistently year after year to be effective.

In addition, individual institutions may set aside specialized reserves for circumstances unique to their mission or environment. For example, the University of Las Vegas at Nevada, which was faced with rapidly growing enrollments, set aside a real estate acquisition reserve to help cope with campus growth (Hignite 2020).

Multiyear Planning

In order to remain financially sustainable, colleges and universities need to think ahead in terms of multiple years rather than the traditional focus of only one year at a time. This provides more lead time to make adjustments by managing vacancies, consolidating functions, streamlining processes, rewriting contracts, and so on.

All universities and colleges engage in some sort of multiyear planning for spending on capital projects, but unfortunately few do this for operating budgets, and of those that do, most do not share the information widely. The purpose of a three- or five-year financial plan is not to predict a given outcome, but rather to help all the stakeholders better understand the relationship between various options and their impact on future finances.

The University of Pennsylvania and the University of Michigan are two examples of universities that do address five-year financial projections as part of the annual budget, but these projections are not published on the websites. The University of Arizona does publish three-year planning parameters on its budget website, but not the outcome of these planning efforts. Other universities may engage in long-term financial planning but have not shared this publicly.

Achieving Buy-in

Managing change in any organization is a challenge, but especially in institutions of higher education, which include a lively collection of well-educated and vocal stakeholders suspicious of higher authority. Buy-in from them is particularly important because a distributed budget system requires a certain degree of trust from leaders of individual colleges, departments, centers, and support units to function effectively.

Most successful transitions to a new distributed budget share certain characteristics in common:

1. A compelling case for change.
2. A commitment to transparency.
3. An extensive consultative process.
4. An appropriate time horizon for adoption.
5. A mechanism for periodic review and adaptation as needed.

Case for Change

In many instances, the precipitating event is some sort of budget crisis that leaves everyone frustrated and open to change. In Ohio State’s case, it was a series of brutal budget cuts targeted by the state of Ohio against higher education in the early 1990s. Campus groups, including faculty, argued they had to make a lot of changes as a result, so why shouldn’t central administration? University officials decided to label the new process “budget restructuring” as opposed to something like RCM in order to make it more consistent with that line of thinking.

The case for change is not complete without a value proposition as to why the proposed changes will make things better. A good example of the case for change for a private university can be found on the University of Pennsylvania RCM website. A good example of the case of change for a public university can be found at the Kent State University RCM website. In both instances, the university leadership stressed how the proposed changes would advance the institution’s academic mission.

Attempts to assess the impact of RCM-type budget systems on revenues have been mixed, but this may reflect the difficulty of measuring something when the applications by various institutions vary so widely. ^[6]

Commitment to Transparency

Achieving buy-in from stakeholders such as faculty and department chairs means university officials need to be transparent in providing them with the financial information they need to make an informed judgment. In some cases that may require the administration to share financial information more widely than they had in the past. I remember when I first arrived at Ohio State, a senior university official told me they did not want to fully open the books, because if the faculty saw how much money was in circulation through the system (even though most of it had restrictions on its use) the faculty would want to spend it all. The problem with that approach is that it produces an information vacuum that feeds conspiracy theories and spreads distrust.

Ohio State officials decided to make opening up the books a feature of the budgeting process. So enhanced information was distributed first to the budget advisory committees described below and then to the campus.

The adoption of the web-based technology opened up additional opportunities for distributed data as well. The RCM websites for Penn and Kent State described above are good examples. Technology has now advanced to allow explanatory videos as well. Good examples can be found at the University of Delaware and Oregon State University budget websites.

Extensive Consultation

The best way to achieve buy-in is to engage stakeholders in the design of the system at an early stage. Most successful transitions do this through some sort of university-wide committee either as part of the established governance process or through a specialized committee created for this purpose. Some have used both.

The key to success in an undertaking such as this is to make sure the committee(s) are involved early, that information is shared on a timely basis, and their concerns are addressed. Membership should include representatives of faculty, deans, and department chairs, staff, senior administrators, and students. Committee members will express a variety of concerns, but some issues in particular surface on most campuses. One is the impact on interdisciplinary work. Another is potential course poaching. And the third is how the libraries are funded.

Higher education coexists with somewhat of an internal contradiction. Colleges and departments are organized along traditional disciplinary lines. But an increasing amount of cutting-edge research and scholarship cuts across traditional disciplinary lines. A good example is biomedical engineering, but it’s certainly not the only one. RCM and its variants didn’t create this problem, but many of those involved in interdisciplinary research worry that a system where the incentives are controlled by traditional units, makes the situation worse, as deans or department chairs are loath to give up revenue or incur additional expense.

There are a number of ways to address this. One is to make sure ground rules on funding of interdisciplinary centers are thoroughly discussed prior to implementation. One option is to treat them as responsibility units as well, so they control their own revenues and expense. Another is for central administration to develop a template on how interdisciplinary grants could share indirect cost revenue or space costs. A third approach is to make interdisciplinary work a priority in distribution of central funding. For example, Ohio State gave interdisciplinary proposals a higher priority for competitive funding in the allocation of Discovery Theme awards, which has resulted in a robust response.

One of the goals of a more distributed system is to give units an incentive to expand enrollment where appropriate. However, it can also be an incentive to do so at the expense of other units, which is not appropriate. The colloquial term for this is “course poaching.” Examples might include math for engineers or writing for business, that siphon off majors from other departments such as mathematics or English. One way to address this is through some form of central review of new course offerings. Many campuses have a faculty committee with wide representation across disciplines that reviews and must approve new course offerings. The key is to conduct these reviews in a timely manner so that it does not become a bottleneck that becomes a barrier to the timely offering of new courses.

Another strategy is for the provost’s office to referee course offerings in popular new areas. For example, instead of letting a free-for-all develop in the growing field of data analytics, Ohio State officials got all the potential stakeholders around the table to hammer out a multidisciplinary data analytics major that included the colleges of Arts and Sciences, Engineering, Business, and Medicine.

Libraries are an emotional issue for many faculty, so how they are funded becomes important to achieving buy-in. Schemes to fund the libraries through some sort of chargebacks based on usage usually meet strong resistance. The more common practice is to fund libraries as a central resource, much the same way as the provost’s office or the registrar.

The specifics of the consultation process will vary depending on the unique character of each campus. Unanimous agreement on all the details is not to be expected, but at least everyone should feel they’ve had an opportunity for input. For example, the University of Florida has laid out an extensive consultation process at its Budget Enhancement Improvement website. It also has a Faculty Senate Budget Council with its own web presence at https://fora.aa.ufl.edu/FacultySenate/Councils/Budget-Council. The Kent State and Oregon State budget websites also lay out extensive consultation structures.

Time Horizon

Timing is everything. Gaining buy-in as described above takes time, and the more time allowed, the more opportunity for a healthy give and take. However, if the ramp-up period takes too long, then stakeholders may lose interest, leadership changes may complicate things, or other crises may intrude. Often overlooked in discussion of a more distributed system is the support structure to allow it to succeed. That includes making sure the budget and accounting information systems provide timely information at the appropriate level of detailed information to the unit level on a timely basis. It also means giving deans, unit fiscal officers, and department chairs the training they need to manage the additional flexibility they have.

The time devoted to discussion and ramp-up to a new system usually runs from three to four years for most institutions. However, there are some significant outliers. In a presentation for National Association of College and University Business Officers, two veterans of the Texas Tech conversion argued that four years was too long because it allowed “too much time for doubt to emerge and positive energy to wane.” They recommended that the process should be limited to no more than two years (Barnes and Clark, n.d.)

In contrast, Ohio State University took eight years, which their officials felt was necessary for successful implementation. They concluded that such a delay was necessary in order to buy time for two supporting efforts to be completed. One was the completion of the strategic plan that the budget model was supposed to support. The other was to allow time to install a new accounting system that would better support the budget model. Both of these were not completed until 2001, but once they were in place, implementation moved quickly, taking only two years.

So in the case of lead time, as in every other aspect of budgeting, each institution needs to adapt its goals and processes to whatever unique factors are appropriate to help ensure success.

Review and Adaptation

The financial environment facing universities is constantly subject to change. Installing a new budget system in a complex institution does not always go according to plan and may produce unintended consequences. Most universities that have done one of these conversions built in some sort of downstream review process to make changes as needed. Two of the first institutions to adopt RCM-like systems are classic examples. The University of Indiana implemented its version of RCM in 1990 and conducted formal reviews at five-year intervals in 1995–96 and 1999–2000. The University of Michigan conducted a review with the appointment of a new provost shortly after Value-Centered Management (VCM) was implemented in 1997. The university adopted major changes in 1999 that included changing the name from VCM to The Budget Model. Additional reviews in 2014 and 2019 concluded that the system was working well but recommended improvements in simplicity and transparency, which were subsequently adopted by the university. ^[7]

Results of other reviews have been posted online by the University of Arizona, University of Florida, Oregon State, and University of Pennsylvania. For example, after a review at the end of three years, Arizona decided to change its system from RCM to “Activity-Informed Budget.” Aside from name changes, one of the more common adjustments was to simplify structure. The classic example is University of Southern California, which reduced its elaborate structure of 100+ cost allocation formulas to four consolidated cost pools (Education Advisory Board, slide 38).

Although the frequency and nature of review vary from institution to institution, it’s fair to say that best practice is to conduct some sort of review every five years or so, to adapt to internal experience or changing external conditions. Many of these reviews have resulted in modifications of the original RCM-type budget model, but there do not appear to be any cases where a university has chosen to revert back to the more traditional system.

Conclusions

In this essay I have examined why many colleges and universities have moved away from traditional incremental budgeting to models that are more directly tied to academic goals, that emphasize long-term financial sustainability, and seek to achieve campus-wide buy-in. Although I have identified examples of best practices associated with these goals, I have also emphasized the importance of each institution’s tailoring its policies and processes to its own unique environment.

What follows is an annotated list of some of the most informative campus websites on non-traditional budgeting.

Alternative Budget Model Websites

Arizona, University of, Activity Informed Budgeting, https://aib.arizona.edu/.

Delaware, University of, Hybrid Budget Model, https://sites.udel.edu/budget/budget-model/.

Florida, University of, University Budget Model Manual, https://cfo.ufl.edu/wp-content/uploads/2020/04/University-Budget-Model-Manual.pdf.

Kent State University, Responsibility Center Management, https://www.kent.edu/budget/responsibility-center-management.

Michigan, University of, Budget Model and System, https://obp.umich.edu/budget/budget-model.

Oregon State University, Budget Model, https://fa.oregonstate.edu/budget/budget-model.

Pennsylvania, University of, RCM at Penn, https://budget.upenn.edu/budgeting-guide/overview/rcm-at-penn/.

Toledo, University of, Budget Modernization, https://www.utoledo.edu/offices/budget/rcm/pdfs/open-forum.pdf.

Additional Information

The librarians at the University of Toledo have put together a comprehensive bibliography on RCM. This can be found online at https://libguides.utoledo.edu/RCM.

The Education Advisory Board has produced a PowerPoint on RCM budget models for Bowling Green in 2015 that the university has shared online at https://www.bgsu.edu/content/dam/BGSU/finance-and-administration/documents/EAB-Budget-Models-BGSU.pdf.

Huron Consulting presented a review of non-traditional budgeting in 2021 for Toledo University at https://www.utoledo.edu/offices/budget/rcm/pdfs/open-forum.pdf.

References

Asher, Herb. 2021. The Gee Years, 2007-2013. Columbus: Ohio State University Press.

Barnes, Mallory and Kyle Clark. (2014?) Responsibility Center Management: The Good, The Bad and The Ugly, Texas Tech University. https://docslib.org/doc/4384263/responsibility-center-management-the-good-the-bad-and-the-ugly.

Hignite, Karla. 2020. “Preserve Your Reserves,” NACUBO Business Officer, March–April 2020. https://www.businessofficermagazine.com/features/preserve-your-reserves/

Jones, Dennis P. 1995. Strategic Budgeting: The Board’s Role in Public Colleges and Universities. AGB Occasional Paper Series. AGB Occasional Paper No. 28. https://eric.ed.gov/?id=ED388146

Education Advisory Board. Bowling Green University. 2021. Optimizing University Budget:  Models, Strategic Lessons for Maximizing Revenue and Mitigating Risk. https://www.bgsu.edu/content/dam/BGSU/finance-and-administration/documents/EAB-Budget-Models-BGSU.pdf.

Jaquette, Ozan, Dennis A. Kramer II, and Bradley R. Curs. 2018. “Growing the Pie? The Effect of Responsibility-Centered Management on Tuition Revenue,” Journal of Higher Education 89(5): 637–676.

Priest, Douglas M., William E. Becker, Don Hossler, and Edward St. John (eds). 2002. Incentive Based Budgeting Systems in Public Universities. Northampton, MA: Edward Elgar Publishing

Whalen, Edward L. 1991. Responsibility Centered Budgeting. Bloomington: Indiana University Press.