Data Driven Decision-Making (3D)
D. Craig Brater, M.D.
Walter J. Daly Professor
Indiana University School of Medicine
June 23, 2004
Background and Rationale
As all of you know, we as a School are embarked Interestingly, when we went through our
on a process that we are calling Data-Driven strategic planning process in 1999, one of the major
Decision Making, or 3D for short. I want to offer recommendations was that we implement mission-
some background so that everyone understands based management. Recall that this strategic
the context of this process and its ultimate aim. planning process was one that sought broad faculty
input and was vetted through each department and
You may have heard of “mission-based various committees of the School. Thus, I have
budgeting.” This is a methodology that has been assumed that it represents the will of the faculty
developed through the AAMC to help schools of and that it represents a guide for the Dean’s office.
medicine make better decisions about how to Many if not most of the recommendations of the
allocate resources. strategic plan have been implemented, but we have
lagged on the 3D project. It is time to rectify that
In its simplest form, mission-based budgeting shortcoming.
entails allocating and distributing specific categories
Why did the faculty make this recommendation?
of resources for specific programs. For example, if a
There was a belief that the
current system is not as
When we went through our strategic effective as it should be and in
planning process in 1999, one of the major particular, it is not sufficiently
recommendations was that we implement robust to meet our future
needs. This makes sense if one
mission-based management…but we have thinks about it.
lagged on the 3D project. It is time to rectify
The current system of
that shortcoming. allocating resources relies on
patterns that were established
state like ours declared that the money it provides so long ago that no one
to the School of Medicine can only be used to remembers their logic or basis. We are a very
support education of medical students, then one different medical school today than we were a
simply adds up the number of hours of medical decade ago, much less compared to when the
student teaching that are done by each department current system evolved. It would be a miracle if a
and the state revenue pie is sliced accordingly. methodology that evolved so long age were relevant
to today’s school in today’s environment. Said in
Obviously nothing is that simple. For example, another fashion, if asked if I am confident that the
how does one enumerate the hours of teaching allocation of resources from the School to each
that a department does? One can easily count Department is fair, based on each department’s
lecture time, but how much preparation time is commitment to and success in each of our missions,
appropriate? For clinical venues where patient I cannot answer in the affirmative.
care is occurring as well as resident teaching, in
a four-hour session, how much time was devoted What scares me is that this may mean that
to students per se? How does one incorporate some departments are not getting the resources
development of new courses or course materials? they need and deserve—in turn, individual faculty
How does one incorporate quality of teaching? may not be getting the resources they need and
Thus, what seems simple on the surface can be rewards that are appropriate. In short, we must
extremely complicated. have systems that meet the needs of a new era;
these systems must be fair, they must promote
One might ask why we should even ponder and reward success and they must be open and
walking down this path in the first place if the transparent.
challenges are so great. Said in another fashion, is
the current system broken such that we need new Without transparency, we risk every individual
methodology; if not, why risk upsetting everyone’s and department thinking that everyone else has
apple cart. a special deal and that the other person’s deal is
Page 2 Data-Driven Decision Making (3D)
better. That kind of environment is not conducive we strive for is informed decision making that
to the collegial, interactive and multidisciplinary is data-driven, participatory and transparent. I
environment that we want to be the essence of our hope you agree with me that there is more than
School. This then, is our call to action. a semantic difference here—in fact, I view 3D as a
methodology that is at the same time an expression
Importantly, I do not see this as “management” of philosophy.
per se, as in mission-based management. What
The Current Situation
Our 3D initiative offers the opportunity to substantially over the last decade. Tuition for in-
explain some of the economics of the School. To state students (about $18,000) is a bit below the
many of you, this area is likely a big black box median for Big Ten schools of medicine and for
about which you have no clue. The way I have Midwest schools of medicine; the same applies
educated myself is to consider the different sources for our out-of-state-students (about $37,000).
of revenues for the School and their intended uses. The total amount of funds that we derive from
Imagine first that you are the Dean and you have a tuition and fees is about $31 million. These data
bank account. The funds that are the “income” for also mean that all other schools in our region have
this account are: been increasing their tuition at about the same rate
that we do. The “cost” of these increases is the
State appropriation: increasing debt burden of our students, now on
average about $100,000. We cannot continue to
This occurs through the state budget- increase our tuition and inflict even greater burdens
making process that occurs every two years. If on our students; thus, this source of funding has
the State increases the budget by X%, then IU limited upside potential.
passes along that percentage to us. However,
our parent university itself has a fee structure
Indirect cost recovery:
that is applied to all Schools. These assessments
are used to support the administrative structure Grant dollars per se go to the individual faculty
of the University as a whole—the financial and in departments; the School receives the so-called
administrative functions, the architect’s office, etc. indirect costs. These dollars are a fixed percentage
of grants that are funded. Our highest indirect cost
As you are likely aware, the State
budget allocation to IU has changed The dollars we have to spend from
little over the past decade; moreover,
expenses that comprise the fee structure the state have changed little in the
have increased—for example, a large past decade…we should not plan our
component is health care costs for
employees. The net result is that the
future around substantial increases
dollars we have to spend from the state in state funding.
have changed little in the past decade.
recovery is for federal grants and is 51.5%.
Our allocation from the State last year (less
How is the indirect cost recovery rate
assessments) was about $50M. Do we have reason
determined? It is the result of a periodic negotiation
for optimism that this will increase meaningfully
between IU and the federal government. The
in the future? I do not. I presume that we will
University tabulates all of the infrastructure costs
retain approximately the same amount of absolute
that are relevant to research and then the federal
dollars, but that we should not plan our future
government negotiates downward. In other words,
around substantial increases in state funding. I
even our highest indirect cost rate does not cover
hope that I am wrong.
our true infrastructure costs.
Tuition and fees: Lower indirect costs are applied to some areas
Students pay tuition and fees and those dollars of research; for example, clinical research, since
remain with the School. Our tuition has increased theoretically this type of research consumes less
Data-Driven Decision Making (3D) Page 3
laboratory space and therefore less infrastructure activity have resulted in anything that might have
expense. These lower rates also do not cover our commercial potential. If so, ARTI is poised to help
true costs. us take advantage of it to the benefit of the School
and the individual.
Do we anticipate increases in this revenue
source? Indeed we do as according to our strategic Clinical revenues:
plan we have set a goal of doubling our research
Last year our overall clinical activity brought in
grants from about $200M to $400M. It should be
about $276 million. This does not include additional
apparent, though, that research does not cover
support from our hospitals in the form of medical
its own costs. One can realize this through the
directorships or outright gifts. Of this, IUMG-SC
discussion of indirect costs above. In addition, there
assesses a 2.5% tax and the School assesses 1%.
are direct costs that grants do not cover because of
The residual remains in each clinical department.
federal salary caps, not paying for administrative
support and not paying for start up of new faculty.
The 1% tax captured by the School has
When one takes all these expenses into account,
traditionally been allocated to the library and to
the basic science departments.
Research is a cost to schools of medicine. The School expects each
department to use its clinical
This represents a clear cross-subsidy that revenues to cross-subsidize
we must address head on. education and research—after
all, that is one of the few
sources for investing in
research costs about 20%. This figure holds for us
these missions and the feature that distinguishes
and for other medical schools.
academic medical centers. How much subsidy
should occur? That is a philosophical question
To reiterate, research is a cost to schools of
but one that cannot be properly addressed in a
medicine. So, why do we do it? We pursue research
thoughtful fashion unless we know how much the
for a number of reasons that include creating a
better learning environment for our students, the
satisfaction of expanding the knowledge base of
At present we have no clue as to the degree
medicine, etc. It is the way that academic medical
of subsidy that is occurring. The need for that
centers define themselves.
data in order to make intelligent decisions is one
of the major drivers for the 3D project. Said in
We have chosen to be a full-fledged academic
another fashion, if some departments are investing
medical center, meaning we are always challenged
heavily in education and research and others are
to find the resources to support our research
not, shouldn’t that influence the distribution of
mission. This represents a clear cross-subsidy that
resources from the Dean’s office? If you agree that
we must acknowledge and address head on. This
also means that the more
research we do, the more it
costs. Again, we must accept
The School expects each department to
this fact and use our collective use its clinical revenues to cross-subsidize
wisdom to figure out how to education and research ... at present we have
make it happen. That is one of
the goals of the 3D project. no clue as to the degree of the subsidy .
Royalties: the answer to that fundamental question is “yes”,
Any intellectual property that derives from then you are tacitly agreeing to the need for the
faculty efforts, whether copyrights, patents, or type of data that we must generate with the 3D
material transfer agreements, results in income project.
that is shared with the inventor, the University
and the School. Last year we received about Another important consideration is our
$365,000. This is an area of considerable upside anticipation of the future of this revenue source. No
potential. To realize it, faculty must be constantly one believes that it is going to increase substantially
asking themselves if their research and scholarly and in fact, I would argue that it may even be a
Page 4 Data-Driven Decision Making (3D)
struggle to maintain these revenues at their current You may be interested in knowing that the
level. Thus, the upside potential for this resource School spends about $2.5 million per year to run our
is low. Development Office. We monitor the “performance”
of this investment on a regular basis and continue
to conclude that we benefit considerably more than
Philanthropy: this activity costs; for example, about $14 million
We have done extremely well in our fund- last year.
raising. In fact, as you have undoubtedly noted
from the above, our other revenue streams have What is the upside potential here? No one
been relatively static. The major source of fuel for knows, but we will certainly try to maximize it,
the growth of the School that has occurred over the because it is one of the few resource areas with
past few years has been philanthropy. potential.
Directing Our Revenue Streams
In summary, the Dean’s office has the above matter what the case, clinical dollars that remain
sources of revenue that comprise my “checkbook”. in departments should be used to cross subsidize
From this checkbook, the School allocates revenues education and research. The question is to what
to each department. An important question to ask degree—again a charge to the 3D process.
is the purpose of each of these revenue streams:
Tuition and fees:
State appropriation: These dollars should clearly only be used to
The State dollars are primarily to support support medical student education.
medical student education
and secondarily research. In Residency program administration,
particular, State dollars are not to
support our clinical mission. Note recruitment and stipends for clinical
also that these funds are not fellows require subsidies that need to be
for the support of residents and
clinical fellows as these trainees
quantified through the 3D process.
are considered to be part of the
clinical enterprise. Indirect cost recovery from grants:
For those who are unaware, stipends for These funds are for the infrastructure expenses
residents are paid by our hospital partners. In of research; therefore, they should only be used to
turn, the hospitals (except the VA) receive funds support the research mission.
from Medicare to pay these stipends. What is not
covered by the hospitals is the administration of Royalties:
residency programs, resident recruitment, and These funds derive from research efforts and
stipends for clinical fellows. Thus, these additional logically should be used to support that mission.
expenses are another area that requires subsidy.
We need to quantify that subsidy through the 3D Philanthropy:
process to make intelligent decisions about what
is the appropriate degree of subsidy that should The vast majority of philanthropy is directed by
occur. the donor for specific purposes. Most often it is to
be used for research, sometimes education (usually
scholarships), and occasionally for a clinical
program. Thus, in most cases the use of these
These funds, that are small in amount owing revenues is not at the discretion of the School.
to our having one of the lowest clinical taxes in
the country, could arguably be used to support It should be clear from the above that some of
the clinical mission. However, since academic the School’s resources should be plowed into the
institutions are expected to use clinical revenues to education mission, others into the research mission,
cross subsidize education and research, one could and virtually none into the clinical enterprise. Do we
also argue use of these funds for other missions. No distribute funds accordingly? No, we do not, and
Data-Driven Decision Making (3D) Page 5
that is why we need the 3D project to help better of funds to departments in a data-driven and
guide us as to fair and thoughtful distribution transparent manner.
Resource Distribution Currently
As noted above, resource distribution currently distribution. For example, among the clinical
is according to historical trends. There are several departments, a few receive about $25,000 per
logical clusters of departments and centers: centers faculty member, several are in the $40,000-$50,000
for medical education that in general have small range and some are much higher (while the average
numbers of faculty and large teaching loads; basic was $39,253).
science departments that have large teaching loads
as well as expectations of successful research Is this distribution according to teaching
programs; two “hybrid” departments (Pathology responsibilities; is it according to research
and Medical Genetics) that have both basic science productivity? The answer is “no.” Should the
responsibilities and also clinical programs; and the distribution be linked in some way to contributions
clinical departments. and commitments in each of our academic missions?
I hope that you agree with me that the answer to
When analyzed within a cluster, I would that question is a resounding “yes.” Therein lies the
challenge anyone to create sense out of the value and purpose of the 3D project.
If we are to somehow link resources with in determining appropriate reward mechanisms
strategic planning and with contributions by for individuals. A more general dataset is what is
different departments and centers, we must then needed at the School level.
develop methods for quantifying attributes in each
of our missions that we value. This needs to be an Gathering these data can only be done at
iterative process and one that does not get bogged the level of each individual, the data for each
down in minutiae. We will learn as we go and department and center being a composite. Doing
constantly refine this approach; to think that we so requires each individual to decide how their time
can get it perfect at the start is unrealistic. is being spent in each mission in which they are
involved as a percent of their overall time. This will
We have different teams that are discussing our not be easy at the start, but it should become easier
various missions. In so doing their task is to decide with time. For a first iteration, we need to start
what things should somewhere. One
be quantified as of my biggest
a first step but The goal is to use our resources wisely fears is that this
to also articulate and fairly. There are no hidden agendas. process is deemed
other attributes to be a “big
that we value but brother” exercise.
that may not be readily quantifiable at present. For I hope that by reading the above, you understand
example, in teaching, it is easy to quantify lecture the rationale and agree with the need. The reality is
and staffing hours. We also value the quality of that it cannot be done at any other level than that
teaching in all venues. If we do not currently of the individual.
have methods for accurately measuring quality,
should we forestall the 3D process until those are As we embark on this project, remember that
in place? The answer is “no,” because we can still the ultimate goal is to use our resources wisely and
generate informative data as we develop additional fairly. There are no hidden agendas. The intention
parameters we deem important. When new data is to be transparent in the process as is reflected in
are available, they can be used to constantly refine the data that have been shared in this dialogue. If
the process. you have questions about any of the information I
have tried to impart, please feel free to contact me
In addition, it is important to realize that the or anyone else in the Dean’s office. Thank you for
highly granular and detailed data are likely best your help in assuring a more solid future for the
used at the department level in annual reviews and School of Medicine.
Page 6 Data-Driven Decision Making (3D)