1996, Number 6 Performance Monitoring and Evaluation TIPS USAID Center for Development Information and Evaluation SELECTING PERFORMANCE INDICATORS What Are Performance Indicators? Simply put, performance indicators are measures that describe how well a program is achieving its objectives. Whereas a results statement identifies what we hope to accomplish, indicators tell us specifically what to measure to determine whether the objective has been achieved. Indicators are usually quantitative measures but may also be qualitative observations. They define how performance will be measured along a scale or dimension, without specifying a particular level of achievement. (Planned levels of achievement -- targets -- are separate from the indicators themselves). USAID operating units have developed hundreds of performance indicators in recent years. Common examples include the dollar value of non-traditional exports, private investment as a percentage of gross domestic product, contraceptive preva- lence rates, child mortality rates, and percentage of eligible voters voting. Why Are Performance Indicators Important? Performance indicators are at the heart of a performance monitoring system -- they define the data to be collected to measure progress and enable actual results achieved over time to be compared with planned results. Thus, they are an indis- pensable management tool for making performance-based decisions about program strategies and activities. Other ways that performance indicators, and the data collected on them, can be used include the following:  to orient and motivate operating unit staff toward achieving results  to communicate USAID achievements to host country counterparts, other partners, and customers and  to report results achieved to USAID's stakeholders, including the U.S. Con- gress, Office of Management, and Budget, and citizens. For What Results Are Performance Indi- cators Required? Reengineering guidance requires operating units to develop performance indicators for all strategic objectives, strategic support objectives, special objectives, and USAID-supported intermediate results (referred to below as SOs and IRs) identified in the results frameworks. Some means should also be developed for gathering information on the results supported by development partners and on the status of critical assumptions, although less rigorous standards apply. Also, SO teams are required to collect data regularly on activity-level inputs, outputs, and processes to ensure they are proceeding as expected and are contributing to relevant IRs and SOs. This implies some thought be given to developing indicators for monitoring progress at the activity level. Steps in Selecting Performance Indicators Selecting appropriate and useful performance indicators is a fairly straightforward process, but requires careful thought, iterative refining, collaboration, and consensus-building. Here are some suggestions. Although presented as discrete steps, in practice some of these can be effectively undertaken simultaneously. Step 1. Clarify the results statements. Good performance indicators start with good results state- ments that people can understand and agree on. Carefully consider the result desired. Review the precise wording and intention of the strategic objective, strategic support objective, special objective, intermediate result, critical assumption, or result supported by partners. What exactly does it say? Avoid overly broad results statements. Sometimes objec- tives and results are so broadly stated it is difficult to identify the right performance indicators. Instead, specify those aspects believed to make the greatest difference to improved performance. For example, rather than using a broad results statement like "improved capacity" of a host country institution, clarify those aspects that program activities emphasize. For example, improved personnel recruitment process, or improved management skills. Be clear about what type of change is implied. What is expected to change -- a situation, a condition, the level of knowledge, an attitude, a behavior? For example, changing a country's law about voting is very different from changing citizens' awareness of their right to vote, which again is different from their voting behavior. Each type of change is measured by different types of indicators. Also, clarify whether the change being sought is an absolute change, a relative change, or no change. --Absolute changes involve the creation or introduction of something new. --Relative changes involve increases, decreases, im- provements, strengthening or weakening in something that currently exists, but at a higher or lower level than is considered optimum. --No change involves the maintenance, protection or preservation of something that is considered fine as is. Be clear about where change should appear. Is change expected to occur among individuals, families, groups, communities, regions? Clearly, a change in the savings rate for an entire nation will be quite different than for a particular sector of the business community. This is known as identifying the "unit of analysis" for the performance indicator. Identify more precisely the specific targets for change. Who or what are the specific targets for the change? For example, if individuals, which individuals? Average citizens or exporters? All exporters or only exporters of non-traditional agricultural products? Study the activities and strategies directed achieving change. Some activities will produce the desired change directly, while other activities will produce the change less directly. For example, activities to develop microenterprises aim to increase employment directly. Activities to reform economic policies may have the same effect, but less directly. Before appropriate indicators can be developed, clarity is needed about the expected relationship between activities and their intended results, in order to understand exactly what changes are reasonable to expect. Step 2. Develop a List of Possible Indicators. There are usually many possible indicators for any desired outcome, but some are more appropriate and useful than others. In selecting indicators, don't settle too quickly on the first that come most conveniently or obviously to mind. A better approach is to start with a list of alternatives, which can then be assessed against a set of selection criteria. To create the initial list of possible indicators, tap the following sources:  internal brainstorming by the strategic objective team  consultations with experts in the substantive program area  experience of other operating units with similar indicators. The key to creating a useful initial list of performance indicators is to be inclusive. That is, view the desired result in all its aspects and from all perspectives. Allow suffi- cient oppor- tu- nity for a free flow of ideas and creativity. Step 3. Assess Each Possible Indicator. Next, assess each possible indicator on the initial list. Experience suggests using seven basic criteria for judging an indicator's appropriateness and utility. These seven criteria are described in the box on page 4. When assessing and comparing possible indicators, it is helpful to use a matrix with the seven criteria arrayed across the top and the candidate indicators listed down the left side. With a simple scoring scale, for example 1-5, rate each candidate indicator against each criterion. These ratings will help give an overall sense of the indicator's relative merit, and help in the selection process. However, apply this approach flexibly and with judgment, because all seven criteria may not be equally important. Step 4. Select the "Best" Performance Indicators. The next step is to narrow the list to the final indicators that will be used in the performance monitoring system. They should be the optimum set that meets the need for management-useful information at a reasonable cost. Be selective. Remember the costs associated with data collection and analysis. Limit the number of indicators used to track each objective or result to a few (two or three). Select only those that represent the most basic and important dimensions of our aims. SEVEN CRITERIA FOR ASSESSING PERFORMANCE INDICATORS 1. DIRECT. A performance indicator should measure as closely as possible the result it is intended to measure. It should not be pegged at a higher or lower level than the result being measured. For example, contraceptive prevalence rate is a direct measure of the result increased use of family planning methods. But number of service providers trained would NOT be a direct measure of the result improved service delivery. Just because people are trained does not necessarily mean they will deliver services better. If using a direct measure is not possible, one or more proxy indicators might be appropriate. For example, sometimes reliable data on direct measures are not available at a frequency that is useful to managers, and proxy indicators are needed to provide timely insight on progress. Proxy measures are indirect measures that are linked to the result by one or more assumptions. For example, in rural areas of Africa it is often very difficult to measure income levels directly. Measures such as percentage of village households with tin roofs (or radios or bicycles) may be a useful, if somewhat rough, proxy. The assumption is that when villagers have higher income they tend to purchase certain goods. If convincing evidence exists that the assumption is sound (for instance, it is based on research or experience elsewhere), then the proxy may be an adequate indicator, albeit second-best to a direct measure. 2. OBJECTIVE. An objective indicator has no ambiguity about what is being measured. That is, there is general agreement over interpretation of the results. It is both unidimensional and operationally precise. To be unidimensional means that it measures only one phenomenon at a time. Avoid trying to combine too much in one indicator, such as measures of both access and use. Operational precision means no ambiguity over what kind of data would be collected for an indicator. For example, while number of successful export firms is ambiguous, something like number of export firms experiencing an annual increase in revenues of at least 5 percent is operationally precise. 3. ADEQUATE. Taken as a group, a performance indicator and its companion indicators should adequately measure the result in question. A frequently asked question is "how many indicators should be used to measure any given result?" The answer depends on a) the complexity of the result being measured, b) the level of resources available for monitoring performance, and c) the amount of information needed to make reasonably confident decisions. For some results that are straightforward and have tried and true measures, one performance indicator may be enough. For example, if the intended result is increased traditional exports, the indicator dollar value of traditional exports per year is probably sufficient. Where no single indicator is sufficient, or where there are benefits to be gained by "triangulation" -- then two or more indicators may be needed. However, avoid using too many indicators. Try to strike a balance between resources available for measuring performance and the amount of information managers need to make reasonably well informed decisions. 4. QUANTITATIVE, WHERE POSSIBLE. Quantitative indicators are numerical (number or percentage of dollar value, tonnage, for example). Qualitative indicators are descriptive observations (an expert opinion of institutional strength, or a descrip- tion of behavior). While quantitative indicators are not necessarily more objective, their numerical precision lends them to more agreement on interpretation of results data, and are thus usually preferable. However, even when effective quantitative indicators are being used, qualitative indicators can supplement the numbers and percentages with a richness of information that brings a program's results to life. 5. DISAGGREGATED, WHERE APPROPRIATE. Disaggregating people-level program results by gender, age, location, or some other dimension is often important from a management or reporting point of view. Experience shows that development activities often require different approaches for different groups and affect those groups in different ways. Disaggregated data help track whether or not specific groups participate in and benefit from activities intended to include them. Therefore, it makes good management sense that performance indicators be sensitive to such differences. 6. PRACTICAL. An indicator is practical if data can be obtained in a timely way and at a reasonable cost. Managers require data that can be collected frequently enough to inform them of progress and influence decisions. USAID operating units should expect to incur reasonable, but not exorbitant, costs for obtaining useful performance information. A rule of thumb, given in the reengineering guidance, is to plan on allocating 3 to 10 percent of total program resources for performance monitoring and evaluation. 7. RELIABLE. A final consideration in choosing performance indicators is whether data of sufficiently reliable quality for confident decision-making can be obtained. But what standards of data quality are needed to be useful? The data that a program manager needs to make reasonably confident decisions about a program is not necessarily the same rigorous standard a social scientist is looking for. For example, a low cost minisurvey may be good enough for a given management need.