Adjusting Program Components For Cost-Effectiveness

The overall cost-effectiveness of a program can be adjusted and improved by looking at cost-effectiveness in segments or program components. This means first finding which parts of the program contribute most to effectiveness and then discovering which of those program components have the lowest cost. Sometimes it is possible to improve cost-effectiveness by enhancing use of these more effective and less expensive components while decreasing use of less effective and more expensive components.

For example, let’s look at County A’s Prevention Program again. Perhaps moving funds from one budget category would be valuable. Perhaps increasing the distribution of the video, reaching more people this way, would increase the effectiveness of this program. Of course, without being certain that the video had the desired (mobilization or other outcome) effect on its viewers, this would not necessarily be more cost-effective.

COMPARING TO ALTERNATIVES

The cost-effectiveness analysis (CEA) frequently compares the costs and outcomes of a program to specific alternatives. The use of comparative CEA has been well established in some fields. For example, in health care, CEA is "a method used to evaluate the outcomes and costs of interventions designed to improve health." (Schoenbaum, M. et al. Cost-Effectiveness Analysis: Overview and Methods. Rand Corporation, 2004.) Cost-effectiveness methodologies are being used to help physicians choose which procedure or medication to use, help insurers decide which services to cover, and help employers decide which benefits (and carriers) to offer.

Similar approaches can be used to assist policy-makers to choose which social programs to support, with the approach being to estimate a particular social program’s effect on the problem it addresses – and then to compare this estimate to another program’s or intervention’s effect on the same problem. The risk of using this type of comparison is that proponents of a particular program might use the technique to justify a program they want funded and manipulate the numbers until a positive benefit-cost ratio is achieved. Opponents of a proposal can do the same. It is thus critical that analysts explicitly characterize their assumptions so that the analysis is transparent. This lends itself to an open process where the issues can be debated on an informed basis. (Cohen, Mark A. Measuring the Costs and Benefits of Crime and Justice. Criminal Justice 2000, Vol. 4, p. 303.)

 RANKING THE COST-EFFECTIVENESS OF OPTIONS

Ranking the cost-effectiveness of various options requires a standardization of measures of costs and outcomes, a process we have yet to develop and implement to a functional level in the domestic violence prevention field. This is because the measures of the effects of domestic violence, and of the effects of domestic violence prevention, are in many ways far more subtle than more straightforward outcomes – such as medical or crime prevention outcomes – for example.

NEED FOR A COMMON METRIC

As Michael Schoenbaum, of the Rand Corporation, in his findings regarding cost-effectiveness writes, "When the same measure of health outcome is used for all interventions, they can be ranked on the basis of their cost-effectiveness ratios. Those with the lowest cost per unit (outcome) are the most efficient ways of improving." (Ibid.) Or as the National Institute of Justice has emphasized, a common metric is required to make cost comparisons. (See box immediately below.)

PERSPECTIVE IS RELEVANT

Comparing alternatives according to their cost-effectiveness depends upon perspective. For example, the societal perspective counts all costs and outcomes, no matter to whom they accrue. Other more narrow perspectives ignore some costs and outcomes, if they are not viewed as being relevant to a particular decision-maker. Narrower approaches to cost-effectiveness analysis (CEA) present the problem of selecting which outcomes are to be measured. (Note the illustration of the narrow versus general approach below).

COST RATIOS

Ultimately, in most cost-benefit or cost-effectiveness analysis, there is a cost ratio calculated, such as in the earlier example which found that the prevention video was viewed by a certain number of students at a cost of $9 per student viewing it. In this case, COSTS ARE DIVIDED BY OUTCOMES this way: 

[(total costs/total outcomes)] =

cost/unit outcome 

[($20,000/2,272 students view video)] =

$8.80/pre student viewing video

or round the $8.80 to

$9 per student viewing video

Or, when dollar-basing the benefits as well as the program costs is possible (and this is not always possible), the formula looks like this:

[(total costs/total outcomes)] =

$ cost/unit outcome in $

This second formula is desirable because:

Simply stating that a prevention video program achieved student video viewing at $9 per student viewing it says nothing about the value of the viewing. Were there a way to say that for each student who viewed the video, the community was saved an average of $100 in intimate partner violence response costs over the next five years, this would establish a dollar-based cost-benefit ratio. 

NOTE:

To calculate the unit outcome in dollars, means to change the “$9 per student viewing video” to the dollar value of having the student view the video . . .Can we prove or even estimate what this will be? Some say yes, and some say no!

Perhaps we can show that for every 100 students who watch this video, the incidence of intimate partner violence is reduced 50% during the five years following the viewing of this video – and that this reduces the cost of this violence from $200,000 to $100,000.

If we reduce this back down to one student, the cost is reduced to 1/100th of that $100,000, or from $2,000 to $1,000.

Therefore in this case we can say that

[($20,000/2,272 students view video)] =

$9/saves $100 in cost of IPV over next five years

 

Spending $9 to save $100 in essence saves $91.

Cost-effective? Yes.

Of course, these statistics are hypothetical. Also, even were these not hypothetical numbers, there is no absolute guarantee that a finding of a high degree of cost-effectiveness guarantees value, let alone long term value.

VALUE IS SUBJECTIVE

In the end, value is, by nature, at its core, (including the value of a prevention program which may be quite difficult to measure), relatively subjective, and each person making the case for domestic violence prevention will necessarily attach value to her or his prevention program.

There are various challenges to cost-benefit analysis (CBA). The choices when doing such analysis are complex, and errors in terms of focus can easily be made. Comparing alternative programs according to their cost-benefits depends upon the comparative or evaluative perspective chosen:

Which benefit outcomes should be measured and compared in order to compare competing programs, and why?

At one end of the cost-benefit approach is a very general perspective, and at the other is a very narrow perspective. While a general perspective may be more reality-based in terms of its being more reflective of the “larger picture,” a narrower perspective is more focused and may allow more precisely applicable information to be generated by a cost-benefit analysis:

	narrow cost-benefit approach		general cost-benefit approach
number of persons in a county who moved out of a domestic violence situation after being educated as to warning signs of domestic violence-related danger at a cost of $104 per person educated by this program this year			change in number of incidents of domestic violence in a state after a state-wide media campaign serving as public education regarding warning signs of domestic violence-related danger at an average cost of $54 per state citizen in one year

There is no best perspective in seeking to know the clearest cost-benefit of a prevention program. However, it is helpful to recognize the risks of looking at the issue too far from one or the other extreme.

GENERAL COST-BENEFIT APPROACHES

The general CBA frequently assumes a very broad, overall, societal perspective, and perhaps rather over ambitiously tries to count all measurable costs and all measurable outcomes, no matter to what or whom they accrue:

The thinking driving this evaluative approach is: “Society is affected, so measure this – all of it -- no matter what it indicates and no matter how the effect being measured occurred!”

Of course what drives this broad analytic approach is the desire to understand what indeed works for the common good; however, it is easy to get lost looking at the forest and not recognize the trees which compose it.

NARROW COST-BENEFIT APPROACHES

On the other hand, some evaluative approaches are too narrow – overly specific. Hence, some of the more narrow perspectives purposefully ignore -- simply choose not to evaluate, and sometimes just omit -- some costs and outcomes, when these are not seen as relevant to a particular analysis or decision-maker. The risk is that sometimes what is relevant is not understood to be relevant, and therefore not measured.

Narrower approaches to CBA present the problem of selecting which outcomes are to be measured. On the other hand, narrower approaches to CBA can render a more useful cost-effectiveness analysis in that the focus is upon a perhaps more precisely measurable linkage between program activities and program outcomes.

COMPARING PROGRAMS BY COST-BENEFIT

Basically, all CEA, and its subset, CBA, intends to look at the cost of a program, and of its components, and then at the pro uo;s outcomes in terms of benefits or lack of benefits. Reviewing a single program in this way is already conducting a cost-effectiveness analysis, while comparing two programs is a comparative CEA/CBA, as seen in the following illustration:

Cost of Program #1	Benefit of Program #1
$250,000  that year	the domestic violence rate among families served reduces by 20% that year which can be given an economic value($12,500 for each 1% reduction)
Cost of Program #2	Benefit of Program #2
$350,000 that year	the domestic violence rate among families served reduces by 50% that year which can be givenan economic value ($7,000 for each 1% reduction)

Two or more programs which intend to have the same outcomes can be compared, looking at the difference in costs of these programs and the differences in their outcomes.

 

Which of the above two programs is more cost-effective according to these measures?

PROGRAM #2.*

*Each dollar spent relates to the cost of 1% reduction in domestic violence.

Please see the Making the Case Worksheet in Chapter 8 for a step-by-step process to determine the benefit of your prevention program.

TIME IS A FACTOR

Time is also a factor in this analysis: how long into the future, how much time after the program being evaluated for its effectiveness concludes is allowed into the calculation of its cost-benefit or cost-effectiveness? How short are short-term outcomes and how long are long-term outcomes?

Does an outcome far into the future have the same value as the same outcome would in the present, now? Is something good happening NOW better NOW than something good happening much later?

Or in other words: Does an outcome far into the future have the same value as the same outcome would in the present time? And if we do find the future benefits of a current prevention program worth measuring, who will fund the measurement of the long-term outcome and who will take these measurements?

For example, a prevention education program is spending $300,000 to directly reach at least 10,000 persons. Each of these people is in a position to reach several hundred other people, who will each reach other people, and so on in a “ripple effect” of information-sharing. Measuring the number of people reached the first year, or during the three years this program is funded, does not measure the long-range impact of this program, and therefore does not provide the full information for totally accurate cost-benefit analysis. Without the capacity to do long-range outcome measuring, and without being able to isolate the effects of a prevention program from other effects which may also have similar outcomes, we cannot measure the full outcome.

IDENTIFICATION OF CAUSALITY

All sound CEA and CBA involves the identification of what is called “causality.” Causality is the linking of a cause and its effect, or in the case of cost-effectiveness and benefit analysis, the program and its cost-effectiveness or benefit. When speaking in terms of the cost-effectiveness of a program, causality is always involved. (See Chapter 6 on the importance of understanding a program’s underlying theory of cause when making the case for that program’s cost-benefit.)

Unless the specific effects being measured are clearly the result of the specific program being analyzed for its cost-effectiveness, then the effects may not be seen as relevant to the evaluation of that program. The effects or benefits may even be mistaken for or confused with the effects or benefits of another program.

Say for example that a media program to prevent new cases of domestic violence begins at the same time another new program is initiated in the same community, this one requiring all people who are seeking a marriage license to take a domestic violence information class. If the rate of new cases of domestic violence drops, it may be difficult to attribute this wonderful development to one specific program. Funders may thus decide that one or the other of these programs is no longer necessary.  Refer again to Chapter 3 on the definition of prevention as being systemic. Here is where the case for both programs, as part of an “integrated community effort” to prevent domestic violence, may be made.