This paper offers a framework for reexamining the set of skills and techniques included in the quantitative and research methods curricula of American graduate planning programs. These offerings, viewed as the supply of skills, are compared to the demand -- skills and techniques used by U. S. planning practitioners. The analysis explores the match between supply, current demand, and skills and techniques practitioners claim they intend to use in the future. Results of the analysis are linked with 1986 work by Contant and Forkenbrock. The quantitative curricular offerings of planning programs are found to be relatively unresponsive to current and future practitioner demand for skills. Directions for possible curriculum changes are suggested.

What quantitative and research methods (QRM) do planning programs teach graduate students? We examine key factors that shape and inform curriculum decisions and propose that practitioner demand for QRM should be a key factor, but not the only one. Using survey data, we assess practitioner demand and compare it to QRM supply -- school-taught techniques -- to inform the debate over quantitative and research curriculum design for the planning profession. This paper is linked to past research that has evaluated QRM demand and supply.

The first section proposes a framework for examining the role practitioner demand should play in curricular decisions. The second section describes the data collection. The third section details the approach taken here to compare the demand and supply of QRM, and presents results of the data analysis. The last section addresses the use of our results in curriculum redesign, and charts some directions for future research.

The preoccupation with content of QRM courses is not unique to planning. Similar dilemmas are faced by public administration and political science programs (see, for example, King, 1982; McCurdy & Cleary, 1984; and, Rodgers and Manrique, 1992). For example, the National Association of Schools of Public Affairs and Administration (NASPAA), which accredits public administration programs, is wrestling with ways in which it should evaluate the quantitative content of curricula, to reflect the changing nature of practice and technology. Within multidisciplinary fields like planning, comprising academic inquiry as well as practice of public decision making, QRM teaching is time-intensive and QRM learning is often difficult. The effort to master quantitative skills sometimes appears to students to be non-commensurate with practical benefits. According to Feldt (1986), quantitative methods are generally unpopular with students and "techniques requiring large amounts of precise data and ... years of analysis are often inappropriate for planning." Therefore, programs need to reexamine periodically which QRM they teach, at what level, and whether they should be core or elective. And, to motivate students to invest effort in acquiring quantitative skills, programs need to give a realistic account of QRM relevance to current and future planning practice.

The menu of QRM courses strives to match each planning program's statement of mission and its view of what planners do and should do. In this sense, curricular decisions are goal-driven. However, constraints discussed below reshape curricula over time, creating the variation we observe across programs.

Planning philosophies impinge upon curricular decisions.

• Quantitative methods gain emphasis when planners are cast as technicians informing political decision makers. This emphasis tends to decline when planners are expected to play community organizer or activist roles. For example, Teitz (1974), adhering to the latter view, called for socially responsive methodology. A curriculum based on this view would assist planners in solving the social problems of the day. Since Teitz rejects the notion that methodology can be value-neutral, the implicit criterion for curricular decisions might be that only methodology that helps shape the planners’ world view should be taught. Today’s users of QRM such as GIS might argue that a method’s usefulness is independent of specific applications, shifting the locus of social responsibility from the methodology to its user.

The organizational make-up of programs contributes greatly to the uniqueness of each program.

• Resources and staff lead the constraints list. Changes in QRM courses impinge on faculty hiring or reassignment, number of contact hours, and computer software and hardware purchases. Needed changes may wait for resources and faculty availability. One concern here is the distinction between a real need for change and initiatives driven by fads, by technology, or by the hiring of faculty specialized in some method. The opposite is a concern with reluctance to implement needed change because faculty expertise in a given technology is not present, or because of faculty inertia.
• Planning programs' history, culture and organizational dynamics evolve consensual equilibria disturbed only at great personal and collective cost; so program changes are difficult and require strong leadership. Even modifications restricted to the QRM curriculum are difficult, as courses grow to rely on each other's content and changes reverberate beyond the QRM area. As a result, the curriculum of the moment may be the result of hard-won victories in the past that are not easily surrendered even if obsolete.

The sum of incremental advances in the technology of computing over the past decade can effect a qualitative change in quantitative reasoning, and in the communication of results to interested parties. The relative ease of performing computations on large data batches, conducting sensitivity tests, and displaying results graphically, can change planning practice. For example, it can alter planners’ relationship to clients, increasing public participation or alienating it further. The rate of adoption of cutting-edge technology varies among programs, despite the fact that QRM have become more accessible with increasingly user-friendly computer software.

• Some programs have current computers and software, and encourage intensive use by students. Others have older technologies whose intricacies discourage student and faculty use. In either case, the QRM curriculum is partly driven by the availability of funds for acquiring the new technology, rather than by what planners believe to be its ideal role.

Pedagogical views shape QRM curriculum design, with resulting variation in teaching approaches and emphases across planning programs.

• To heighten the students' awareness of QRM relevance, discussants at the 1985 ACSP conference suggested setting quantitative methods in their planning context, and including implementation issues (Forkenbrock, 1986). Underlying such calls is a pedagogical model holding that students learn more readily through induction -- by seeing the immediate applicability of a technique to a real problem. However, only precious few QRM, at the lower end of complexity and usefulness, can be acquired inductively at any but a superficial level. There may not be enough time within one course for dealing with a complex, real problem, and mastering a complex technique too. In addition, students not previously exposed to quantitative reasoning and computers have difficulty studying statistics and computer applications simultaneously, and may confuse computer interface troubles with the challenges of quantitative procedures.
• Schuster (1986) proposed organizing a planning program’s core around five filters: historical, institutional, political/economic, quantitative argument, and literate. QRM in his view should be taught as components of the quantitative argument filter, not as an end in themselves, as they may appear to students when taught outside the planning context. Schuster cautioned against a simple correspondence between these filters and actual courses, and recommended integrating the core and linking it to electives. Combining in one course real problem solving with QRM learning, however, is quite challenging. The difficulties may have led to the predominant form of QRM teaching, which focuses on methods and techniques illustrated through simplified examples of the real situations to which they apply. Some programs follow specialized QRM courses with integrative ones that require real problem solving, including structuring research questions and selection of quantitative tools, possible only after students have acquired a tool kit. Other programs do not actively incorporate QRM in substantive courses, hence students' perception that some of their most difficult courses do not have any real use.

The level of competency to be attained in QRM courses bridges between student and practitioner considerations, in that it should be driven to an extent by practice. This complex curricular decision combines practitioner and student ability constraints with planning and pedagogical philosophies and with the limits on the number of hours required for graduation. The typology of competency levels we propose to use here, described below, is consistent with both Contant & Forkenbrock (1986) and Prosperi (1986).

• Graduates should enter practice equipped with QRM matching current needs to a great extent. For example, if planning positions require considerable technical writing, graduates should have competency in this skill -- the ability to do it independently, taking responsibility for data gathering, input, computational aspects and results interpretation. This is roughly equivalent to Contant & Forkenbrock's criteria of critical and analytical thinking, rigor in procedure, and depth.
• If statistical procedures are typically handled by a departmental statistician or a consultant, then graduates may not need mastery, but should have a working familiarity with the techniques -- know when to ask for the procedures, understand its computational nature, judge the appropriateness of its use, interpret the results and factor them into decisions. This is equivalent to "greater familiarity", from earlier work cited.
• Finally, some techniques mentioned in published research may be relevant to practice; then practitioners may only need awareness of the nature of those techniques, of their role in analysis, of the data input and nature of results. This competency level is similar to Contant & Forkenbrok's basic introduction category of competence.

Students' demographic and professional characteristics may affect the content and number of QRM courses offered.

• Today's programs serve at times a relatively older group of students reentering school after some years of practice. Their interests and needs are largely driven by their current job, against which they tend to assess the usefulness of quantitative techniques. As we argued, however, current practice may not have caught up with cutting-edge or highly technical skills.
• Students' initial skill level drives the level of competency attainable in class. Programs dependent on local draw need to be especially responsive to the students' incoming skill level, which is often shaped by the region's public school systems. Only relatively small programs can respond to classroom skill diversity by tailoring course content to individual students. When this is not feasible, programs may allow students to choose paths or early specializations, and may relegate more advanced QRM to elective status. In such programs, unless "leveling" courses are provided to fill gaps, QRM have to be taught at the modal incoming skill level, which often enables only awareness, or at best familiarity with key QRM. Leveling courses may become an enrollment disincentive however, as they add credits, cost and time. Pressure to eliminate the more technical, learning-intensive QRM may ensue.

Current practitioner needs are mirrored by the job market. Practitioners' future needs depend on the length of the horizon considered: incremental short-range changes do get reflected in the job market, while more drastic long-range changes typically become evident only in retrospect.

• Job descriptions and hiring patterns contribute to programs' perceptions of skills in demand that enhance graduates' placement chances. In turn, demand for skills and competency levels is somewhat shaped by the QRM knowledge of graduates become practicing planners. Recent graduates form but a fraction of the practicing body, so in general practitioners do not have competency in skills at the "cutting edge" and are, therefore, less likely to feel a need for such skills. So curricula driven exclusively by practitioner demand may shortchange future practice by teaching yesterday's techniques. Curricula ignoring current demand are shortsighted too, impairing graduates' ability to compete in the job market.
• To respond to current demand for QRM, planning programs may produce a well-rounded, adaptable graduate, with familiarity- and awareness-level skills in a wide range of QRM, who will reach mastery on the job and who will have access to numerous market opportunities, but will compete with both specialized and unspecialized graduates. Another strategy is to produce graduates with mastery in some highly technical skill, who will compete in a narrower market but only against other specialized graduates. Geographic information systems (GIS) is such a niche. The first strategy may be robust to changes in the long run, but the second strategy, if flexible, can take advantage of short-run job market needs.

Of the curricular decision factors described above, planning philosophy, technology and practitioner demand have, arguably, a more consistent effect across programs, while institutional constraints, pedagogical views and the student body characteristics tend to be program-specific. If so, we should expect similarity in the kinds of QRM taught at planning programs, and variation in competencytime devoted to particular techniques and in competency levels sought.

The proposed curriculum decision framework highlights the fact that current practitioner demand is not the sole factor driving QRM menu choices, even if it is prominent among programs' concerns. For example, current demand may lag behind current supply because recent graduates exposed to cutting-edge techniques are typically not in the organizational positions where QRM choices are made. Further, QRM use may necessitate adoption of new technology requiring investments not always within the immediate reach of planning agencies. Technological lags are currently experienced, for instance, in the adoption of computer-aided design and geographic information systems, which typically require considerable investment in hardware and software, extensive staff training, and possibly even some organizational and procedural changes.

These examples suggest it might be a strategic mistake for planning programs to aim at balancing supply with current practitioner demand, especially in the area of highly technical skills relying on newer technologies. However, it might be equally unwise to neglect current demand for QRM and to fail to equip graduates with skills that will assist them in the job market. QRM supply and demand feed each other in ways that do not allow simple inference from data at one point in time. Therefore, we do not expect a one-time comparison of QRM supply with demand to show balance, but supply and demand changes over time should indicate whether, and to what extent, programs respond to various goals and constraints, including practitioner demand. Also, while not conclusive for normative purposes, practitioner demand should reveal whether planning programs are bringing QRM advances to bear on planning practice. Planning programs may want to consider both current and future demand in their QRM choices, and to keep records that enable the capture of changes in time.

Two notable evaluations of the content of QRM planning curricula and their consonance with practitioner needs were performed in 1975 (Schon et al., 1976) and 1985 (Contant & Forkenbrock, 1986). In designing the present research, we set considerable weight on comparability of our results with the previous ones. The list of QRM we used to survey planning programs and practitioners incorporates the items covered by Contant & Forkenbrock and by the AICP planner survey of 1992, to help discern trends in some specific skills and techniques programs teach and planners use. For example, if planning programs do react to practitioner demand, we should expect changes from 1986 to 1992 in QRM taught (supply) to be in the direction that brings them closer to balance with the practitioner demand as assessed in 1986. If schools had an impact on practice, we should observe 1992 practitioner demand moving in the direction of 1986 supply of QRM.

Contant & Forkenbrock (1986) inquired about 27 QRM areas. They asked 65of 95 programs about the techniques they taught (22 QRM on average); and, they asked 69 planning directors from the AICP roster to what extent planners in their agency should possess knowledge about each listed technique. In addition to providing useful information about planning programs and about the planning practice of the time, this project compared QRM supply (techniques taught in programs) to demand (techniques found valuable by planners).

The comparison of current QRM offerings in planning programs with the Contant-Forkenbrock survey is limited by the lack of an exhaustive, non-redundant list of skills taught. Also challenging is establishment of a direct correspondence between courses taught and methods used by the practitioners. The lack of fully compatible longitudinal data also poses difficulties. Therefore, the value of our research lies as much in the results as in its structuring of the debate and the challenge to devise better tools for data collection to enable more accurate analyses in the future.

How should a planning program evaluate its QRM curriculum and change it if necessary? Our research evolved from an evaluation and change initiative stemming from faculty and student dissatisfaction with some aspects of the QRM courses at the Levin College of Urban Affairs (LCUA) at Cleveland State University. The LCUA effort yielded not only needed curricular changes, but also insights into the decision process of curricular design in general, and data on supply and demand of QRM for evaluating and changing QRM curricula in other programs and for debates over what planning students should learn. The data sets used to explore the match between demand and supply of QRM are described in the next section.

The data collection for this project entailed a mail survey of planning programs in the U. S. and Canada, and a telephone survey of AICP-listed practicing planners in non-academic positions, on the teaching/use of 53 listed quantitative research methods (QRM).

The demand for QRM is represented by the use patterns among planning practitioners. Telephone interviews were conducted with 106 practicing planners during Summer and Fall 1992. We used stratified random sampling (by state) of the 1990 AICP directory, from which we excluded all members with academic addresses.

While the surveys and interviews elicited a wealth of information related to QRM teaching and patterns of use, the key data for the supply - demand analysis are the frequency of teaching, of practitioner use, and practitioner ratings of the relative importance of QRMs. We broadly defined QRM to include project methodology (e.g., linear programming, logit models), project design (research design, sampling), project implementation (surveys, data management), and analysis of results (descriptive and inferential statistics).

We compiled a QRM list consistent with Contant & Forkenbrock (1986) and AICP (1992), to assess change in use of techniques since 1986 and enable future analyses with 1992 as a baseline. As a result, our list contains several non-mutually exclusive QRM entries, particularly for statistics, which may limit the extent to which our results can be interpreted for those categories. Our sampling frames differ from earlier work.: Contant & Forkenbrock (1986) queried planning department chairs and planning directors about 27 QRM, while our surveys on 53 QRM were addressed to the faculty responsible for teaching QRM (at the discretion of the program chairs) and to AICP planners at several organizational levels (from the same pools of American planning schools and the AICP directory respectively).

At responding graduate schools of planning the most frequently taught skills were descriptive statistics (at 95.1% of the respondents' programs), population projections (87.8%), and regression analysis (85.4%). The least frequently taught skills are multiattribute utility theory (4.8%), stochastic processes (4.8%), non-linear programming (9.5%) and queuing theory (9.8%).

Planning practitioners most frequently use budget preparation (98.7%). Least utilized are logit/probit models (39.7%), shift-share analysis (41.7%) and multiattribute utility theory (41.5%). The most highly rated skills (reflecting current time demands) were data collection (3.4), budget preparation (3.4) and scheduling (3.0). Future practitioner plans (for skills not already in use) are most likely to include GIS, data collection and algebra. The latter may indicate recognition of a need to brush up on basic math skills.

The frequency of teaching or use of each QRM provides an information base for comparison of program supply with practitioner demand for quantitative and research skills and for curriculum design. In the next section we analyze demand and supply trends using Contant and Forkenbrock's (1986) results as a baseline.

Our analysis has three time-related components. The first links our current supply frequency data to Contant and Forkenbrock (1986) demand results; the second, the pattern of gaps between current demand and supply of QRM, uses current frequency of QRM teaching/use; the third consists of our findings on a supply-demand comparison that factors in future practitioner demand for QRM. Our focus is curriculum redesign, so throughout the analysis we take the programs' perspective, examining the degree of balance with practitioner demand and identifying over- and under-teaching of QRM.

• Current-balance scenario: From a short-sighted, static point of view, the goal driving curricular choices of QRM might be to respond to what practitioners need now.

• Forward-looking scenario: From a longer-range, dynamic perspective, lower-complexity QRM with relatively stable demand (such as descriptive statistics) would be supplied at the level demanded by practitioners, while more complex, cutting-edge skills would be taught above their demand level, to foster and satisfy future QRM demand.

Interestingly, not one skill listed in 1986 registered an increase in supply. Shifts from 1986 to 1992 reflects a marked drop in supply: in 1992, programs were teaching 15 of 25 QRM less frequently than in 1986. The bottom row of the table (low supply, QRM taught by 0 to 40% of planning programs) was empty in 1986, but in 1992 it had acquired 11 techniques.

Of the 15 drops in supply, three are moves toward the diagonal (econometric modeling, gravity modeling and present value analysis), as would be expected under a current-balance scenario, were it not for the fact that these are not low-complexity skills. Of the three techniques that showed less demand by practitioners in 1992 (multivariate statistics, regression analysis and trip generation modeling), only one, trip generation modelling, dropped considerably in supply as would be expected as a result of the current-balance scenario -- more complex QRM not currently in demand fall in disfavor over time, as there is no influx of practitioners trained in their use. Four techniques migrated down from balanced cell 5 (medium supply and demand) into low supply, but two techniques moved down into cell 5: gravity modeling and present value analysis. Balanced cell 3 (highest demand, highest supply in 1986) gained nothing but lost supply in 5 techniques. Cell 8 (low supply, medium demand) gained 6 QRM in 1992, at the expense of both medium and high supply cells. 10 of the 15 supply decreases were high-to-medium, or medium-to-low supply.

The predominance of supply reduction, apparently unaccompanied by a parallel change in demand, is only surprising if we believe practitioner demand is the sole driving force behind curricular changes. We argued in the first section, however, that demand is but one of many factors influencing QRM decisions. Our results suggest practitioner demand has not even been a key factor. On the other hand, decline is consistent with part of Teitz’s (1974, p. 97) view that "the ultimate fate of most methods ... will be decline and replacement."

The drop in QRM supply from 1986 to 1992 may reflect planning programs’ response to student demand. Students may prefer fewer total credit hours towards a degree, or courses in other areas of planning. Under pressure to maintain or reduce the number of required credit hours -- to compete with weekend or executive Masters programs -- programs may make trade-offs among disciplines, which are made easier by the fact that, at least anecdotally, QRM courses are rarely student favorites. The cumulative character of QRM learning exacerbates the difficulty of teaching them, as planning programs cannot choose, for example, to teach time series analysis and skip regression, or to teach decision analysis without coursework in probability theory and statistics. A survey of planning students’ attitudes towards QRM would help test these hypotheses.

Results of the previous analysis suggest that on average practitioner demand plays a limited role, if any, in curricular decisions. If planning programs wanted to incorporate practitioner demand in their QRM curricula according to either the current-balance or the forward-looking scenario, they would have to close some of the supply-demand gaps. To assess these gaps, we extended Contant and Forkenbrock's study by considering both the frequency of QRM teaching/use, and the importance practitioners attach to them.

Practitioners rated current QRM importance on a 1-5 scale, relatively to the time spent using a skill; those who were not currently using a skill rated expected future usage on the same 1 - 5 scale. Opinions about future use of a QRM are less informed than those referring to actual use, but nevertheless practitioners observe and evaluate products of QRM they do not use. They read reports, and take courses and workshops which expose them to QRM not available at their office. And, software demonstrations can also effectively inform and create demand for future use. For example, practitioners are acquainted with GIS products even if their organization has not acquired the software.

These results suggest that planning programs mostly do not teach the QRM practitioners use and deem important currently, and not even what they expect to use in the future. However, the pattern of results gives no license to interpret this apparent mismatch as a failure of planning programs to become relevant to practice. Rather, it buttresses the more complex view of QRM curricular decisions proposed in the first section, and points to several directions for needed research, discussed in the last sections.

Planning programs that decide to incorporate demand levels in their supply decisions need to select normatively the weights to be attached to current and future practitioner demand. Not surprisingly, to a large extent this choice hinges on the whole set of factors discussed in the first section: practitioner demand weights correspond to strategies the planning programs need to adopt in order to serve their students and communities.

• For example, one program may decide to respond mainly to current conditions, to ensure graduates are equipped with currently needed skills. That strategy might serve well programs whose graduates are not highly mobile and need to tailor themselves to the jobs available locally.
• Another program may adopt a long-range strategy, setting a lot of weight on expressed future demand. This strategy serves programs looking to ensure future demand for its graduates and for its services, through a stream of graduates who will promote and demand use of newer QRM and who may return to the program for expert advice.
• Yet another program may seek to shape the future of planning by disregarding practitioner demand and by developing and teaching QRM predicted to fit perceived future needs and technologies. This strategy serves programs with national reputation and with resources to adopt or even develop cutting-edge technologies to be tested in the market of the future.

Our survey results can help a planning program in structuring the debate when evaluating the adequacy of its QRM sequence. We tested their use in 1992 at the Levin College of Urban Affairs (LCUA) at Cleveland State University, when we were asked to study the adequacy of the College's QRM sequence, and to propose changes. Any change initiative has to be accompanied by a process enabling implementation by including, informing, and fostering consensus. Our mandate did not encompass debate over the underlying planning philosophy or major technology shifts (beyond software). So the information base for designing and negotiating change at LCUA involved mainly establishing the degree of congruence between the QRM supplied by LCUA courses and current practitioner demand for QRM.

LCUA is a commuter school catering largely to in-service students. It has a new, unaccredited planning Master degree, and 11 planning faculty. LCUA also offers Master programs in Urban Studies and Public Administration, as well as an undergraduate degree and a doctorate in Urban Studies. Total graduate student enrollment, including full and part time students, is about 200. All Master programs at LCUA share the same core sequence of two quantitative methods courses.

The bimodal distribution of incoming math, writing and quantitative reasoning skill levels tends to frustrate both students and faculty. A quarter system and the typical scheduling of QRM courses in one weekly night session of 3.5 hours combine to leave a rather narrow window of student heightened awareness. These factors, together with a widespread student view that the two QRM courses are a test of endurance, led to a perceived need for a QRM sequence reassessment.

We examined the supply-demand gap profiles and designed several alternatives consistent with our goals and criteria, differing mainly in the relative weighting of supply, current and future demand, and in the handling of our students' bimodal entry skill levels. The alternative eventually implemented reflects the realities of institutional and technological constraints as well as philosophical differences expected among multidisciplinary faculty. The gap analysis helped this task in two ways: it allowed testing of alternatives based on different demand and supply weights corresponding to different faculty views; and, it contributed to the communication of results to other faculty who found the information transparent and supportive of debate. Clearly, change implementation hinges critically on consensus, and using the supply-demand gap tool enabled a measure of consensus at LCUA.

This research extends the information available to planning programs for making curricular decisions in the area of quantitative and research methods. Our approach consisted of assessments and comparisons of supply -- QRM courses taught by planning programs -- with demand -- QRM currently used or intended for future use by practicing planners. We linked our results to previous studies to identify QRM demand and supply trends.

The proposed curriculum decision framework highlights the fact that practitioner demand, while prominent among programs' concerns, is not the sole factor driving QRM menu choices. Therefore, the comparison of QRM supply with demand at one point in time should not have been expected to show balance. However, supply and demand patterns in time should reveal whether practitioner demand is factored in QRM curricular decisions, and whether planning programs are successful in bringing QRM advances to bear on planning practice.

Our demand-supply analysis indicated the expected imbalance between supply and demand. Only 43% of QRM were supplied in balance with 1992 demand, and even fewer, 26%, balanced supply with combined present and future demand. Interestingly, the undertaught QRM were of the non-cumulative type, in that typically their teaching does not have to build on previous coursework (budget preparation, issues analysis and scheduling). In contrast, the QRM in the group that balanced demand and supply tended to be cumulative, and overtaught QRM were typically those serving as prerequisites to the cumulative QRM.

The comparison with the Contant and Forkenbrock study (1986) revealed a rather surprising lack of responsiveness among planning programs over time to practitioner demand for QRM. The observed diminishing trend in QRM contradicts both our proposed "current-demand" and "forward-looking" scenarios: no attempt to either balance current demand or create future demand for cutting-edge methods was apparent. Since the diminishing supply trend is not accompanied by a corresponding drop in practitioner demand, we are led to conclude that planning programs do not seem to teach what practitioners practice, and not even what practitioners should practice.

The curriculum design framework was proposed to assist in the use of our supply-demand results in redesigning QRM offerings in planning programs. The framework suggests some areas needing further investigation. For example, student demographic trends and a survey of students' interests, experience, and incoming skill levels may shed light on the observed drop in QRM supply. A survey of available technology and its use at schools and planning offices may also help explain the observed QRM supply and demand patterns. Content analysis of planning programs' syllabi may inform on the nature of variation in offerings among schools, as well as on the competency levels achieved across programs, which may have a serious effect on current demand for QRM.

The effect of available textbooks and teaching materials on the content and pedagogy used in QRM courses should not be underestimated, although it was not explicitly included in the present framework. Ideally, teaching materials can be devised and tailored to the goals and content of any program. In practice, however, time and resource constraints as well as the nature of academic rewards lead to second-best approaches. Courses often rely on available textbooks, which go against Schuster’s caution to "resist falling into the trap of teaching statistics" (Schuster, 1989). Current textbooks and their impact on QRM teaching, whose assessment exceeded the scope of this study, should be a part of future research agendas.

Several hypotheses were advanced in the first section, in presenting the theoretical framework for curricular decisions. They suggest the need to pursue research that would elucidate the decreasing trend in QRM supply by planning programs. For example, to what extent is there pressure to contain or reduce the number of credit hours required for a degree? Do students dislike the study of QRM or is this a symptom of dissatisfaction with teaching approaches in this area? Does the decrease in QRM supply match a decrease in the degree of preparedness of incoming students, or does it correspond to a decrease in the number of new faculty ready and willing to teach QRM? Finally, is there a trend toward the specialization of a few (consultants) that supersedes the need to teach QRM to the many?

As was briefly mentioned in the first section, planning is not unique in its QRM teaching and use needs. Public administration and political science, programs, for example, face much the same challenges in designing QRM courses relevant to practice. The similarities and differences among these academic/practice fields in teaching and use of QRM need to be examined, to establish to what extent approaches and materials are transferable among these disciplines. This need is heightened by the fact that almost two thirds of graduate planning programs sampled were housed with other disciplines.

The periodic assessment of QRM supply and demand is necessary, if planning programs are to serve well the planning profession in all its diversity. However, the difficulties encountered in this study should serve to caution those who will offer to take stock in the future. Comparability with past studies is important, but challenging and difficult to maintain, as the lists of QRM taught or in use tend to change. One avenue for future exploration is the compilation of a QRM list that is based at least in part on practitioner responses to an open-ended interview that would allow the researcher to identify and eliminate problems of mismatch between QRM names and content, as well as problems of redundancy and lack of relevance to practice. Such an investigation might also result in a list that reflects the subordinate, but necessary role of some QRM (such as probability theory) which would otherwise persist in the overtaught category simply because they are building blocks for the more complex QRM. It should be noted, however, that the difficulties of QRM demand and supply assessment go beyond survey design intricacies.

Surveys are afflicted by declining response rates for several reasons: they are time-consuming, individually unrewarding, and proliferating in all walks of life. Electronic communication may alleviate in the future some of the tedium associated with managing, responding to, and compiling results of surveys. Until such a time, the exercise will often be taxing on all involved. In addition, planning programs are very diverse in content, and practitioners span an increasing array of tasks (Glasmeier & Kahn, 1989). Therefore, averages become less instructive and more may be gleaned from the study of individual program and planner profiles. Future initiatives to assess supply and demand of QRM may need to use a combination of surveys, and interviews to capture philosophical differences, institutional constraints and strengths, and emerging planner roles. A component of this system which received the least attention in the past is the planning student, who does, and should, affect planning programs’ content.

We thank LCUA's Dean David Sweet who funded this research, and our colleague Michael Spicer who provided the impetus for it. We are grateful to all faculty and planners who took the time to respond to our surveys. We had the benefit of suggestions from many of our Levin College colleagues. Special thanks to William Bowen and Lawrence Keller who helped compile the list of skills central to our work. We valued the constructive suggestions of our three anonymous reviewers. Any errors of fact or interpretation are, of course, ours.