Transcript Slide 1
Distinguishing Inputs, Outputs, Outcomes and Impacts www.summerinstitute.eu Important points to start with ■ Decision making ■ Policy analysis ■ Limited rationality ■ Incentives ■ Process and learning Contents ■ Outputs & inputs ■ Outcomes ■ Indicators ■ Logic model ■ Budgeting & PI ■ Future Contents ■ Outputs & inputs ■ Outcomes ■ Indicators ■ Logic model ■ Budgeting & PI ■ Future Outputs - definition ■ Services or goods which are provided by ministries and other state organizations to external stakeholders / beneficiaries (directly or through other organizations) ■ External stakeholders include: general public, citizens, businesses, NGOs, media and other state bodies including the minister, other ministries and state organizations, and the parliament Examples ■ Provision of prosthetic devices to disabled ■ Provision of schooling ■ Health centre services ■ Analysis and policy advice ■ Laws ■ Inspections ■ Training Outputs should not be ■ Activities ■ Processes ■ Inter-mediate output ■ Capacity building initiatives Example: output – schooling ■ Number of graduates from school ■ Activity ■ Process - teaching recruitment of teachers ■ IMO – Number of enrolled chidlren ■ Capacity building – increasing professional qualification of teachers Consumption or provision? ■ Economic theory view: the delivery of service takes place between producer and consumer, and therefore transaction is complete, when service is used (output consumed) ■ Theatre plays ■ Patients checked ■ Public administration view: for many outputs it does not mater whether they are consumed or not – the service should be there in case it is needed ■ Emergency services ■ Prison services Output characteristics ■ Volume or quantity of provision ■ Timeliness in provision ■ Quality of provision ■ Satisfaction - surveys and evaluation ■ Comparison – to others or standard ■ Coverage, meting of demand and accessibility ■ Equity Example: output – provision of prosthetic devices ■ Volume – number of disabled people served ■ Timeliness – service provided in specific time – for example, one week after request ■ Quality – satisfaction of users of devices ■ Coverage – % of people in need of devices served ■ Equity – people with access difficulties served at home or transported to the service point Selectivity ■ Not all output characteristics can/should be used for all outputs – be selective ■ Focus on key outputs for external users ■ For all other outputs – have at least the volume Aspects to consider ■ Control of provision ■ Other government interventions ■ Measurability ■ Attribution ■ Users ■ Incentives & gaming Control of provision ■ Provision of a single output is controlled (directly or indirectly) by government organization ■ But not its effectiveness (achieving desired policy objective). Many government outputs make sense only when used together with other outputs ■ Road safety and reduction of fatal accidents depends on a group of outputs – quality of road, speed control, technical state of cars, use of seat belts. Control of provision (contd.) Many organizations Policy Budgeting One or more organizations Delivery One organization Other government interventions ■ To implement policies governments provide not only goods and services, but also: ■ regulation ■ capital investments ■ social and various other benefits & compensations ■ subsidies ■ If these are not outputs, how do they relate to them? Other government interventions (contd.) ■ It can be helpful to think of three broad categories ■ goods and services = outputs ■ administered items ■ capital investment ■ mandatory benefits & compensations ■ subsidies and other payments ■ general regulation ■ But remember – most of administred items and general regulation require outputs Example – road safety Outputs Current services Investment Mandatory benefits Subsidies and other payments Road rehabilitation Policing & speed control Driver education ■ In budgets, some countries separate outputs / investment / financing in separate programs while others do not (but separate using economic classification) Measurability ■ Some outputs are harder to measure than others due to Measurement difficulty (output) Ambiguity the nature of organizations providing them Embassies Army Road police Residential care Job Counselling Policy making ministries Routine (process) Training institutions Unemployment agencies Attribution ■ Outputs mean something if they can be attributed to some objective (outcome) – they contribute to achieving of objective, e.g. Change in outcome ■ Generalized attribution: re-training of unemployed contributing to reduction of unemployment ■ Specific: re-training of unemployed contributing to employability of the trained person during the next year after training Difficulties in establishing attribution Clear attribution: – Broken arm – Medical service – Fixed arm – Services – Survival Less clear attribution: – Cancer – Earlier diagnosis? – Better treatment? – Healthier life style? – Beneficial effects of affluence? Users ■ Different PI users and decision making stages might have different information needs (and therefore require different formats of presentation of PI) Parliament Government Ministry of Finance Minister Permanent Secretary Unit Manager Employee Output focus ■ Using of PI means analysis and learning rather than money and / or promotion Principal – Agent Problem ■ Principals are “owners” and “bosses”, but they lack time and expertise to do everything ■ Therefore principals hire agents, who have time and expertise, but are self interested Medieval Business Modern Principal King Shareholder People Agent Lords CEO Bureaucrats Delegated authority Tax collection Profit making Service delivery Principal – Agent Problem (contd.) ■ Agents have one strong advantage over principals information ■ This creates problem called “gaming”, which is based on two factors: ■ There is information asymmetry between the principal and agent; and ■ There are wrong incentives for disclosing or hiding information Gaming & incentives ■ Bevan and Hood: "reactive subversion such as 'hitting the target and missing the point' and / or reducing performance where targets do not apply" ■ Gaming = manipulation with information that might affect service delivery ■ Incentives = factors (financial or non-financial) that provide motives for a particular course of action, or count as a reason for preferring one choice to the alternatives ■ Targets are the most common causes for perverse incentives Gaming & incentives (contd.) Gaming potential Information asymmetry High Low Loose Relationship with decision-making Tight Impact of gaming TARGET Strategy A Arrival on time at final destination Strategy B “False “Honest Fool” Champion” or or no service manipulation manipulation Strategy C Strategy D “Smart Liar” or manipulation of data “Saboteur” manipulation of service and data Impact of gaming (contd.) ■ Gaming can result in: ■ Cherry picking – handling of easy cases, patients etc. at the expense of difficult ones ■ Ratchet effect – slowing down today to be on time tomorrow ■ Distortions – improving areas where measurement is dome while neglecting other areas ■ Threshold effect – focusing on 2nd best and forgetting about the best and worst, as best will be achieved anyway while worst require too much of an effort Dealing with gaming ■ Learning and data improvements ■ Avoiding strong negative incentives ■ Clarify use of PI in advance ■ “Loose” rather than “tight” connection to decisio making ■ Provide feedback on data relevance and quality ■ Reward those with good PI ■ Do not overload with data – use the key measures Dealing with gaming (contd.) ■ Dealing with principal – agent problem at all levels ■ Use of “auditors” ■ National Audit Office ■ Internal audit ■ Encourage ownership at production level Summarizing about outputs ■ Goods and services ■ Basic element in performance information system ■ Can be measured in different ways, but there are difficulties ■ Thee are three most common difficulties: ■ Measurability ■ Attribution to objectives (outcomes) ■ Scope for manipulation ■ Loose link between output data and decision making generally is a better incentive to avoid gaming Inputs - definition ■ Inputs are resources used in the delivery of goods and services ■ Labour and other means of service provision (facilities, computers, teachers, books etc.) ■ Money is not an input. It s a cost of input Inputs – output combination More outputs V ? Less inputs More inputs ? X Less outputs Improving performance More outputs Less inputs Production possibility frontier More inputs Less outputs Contents ■ Outputs & inputs ■ Outcomes ■ Indicators ■ Logic model ■ Budgeting & PI ■ Future Outcomes - definition ■ The effects which government is trying to achieve for the public. They are often expressed as improvement in the living conditions of people or favourable changes that contribute to these changes ■ Outcomes are changes in the economic, physical, social and cultural environments which the state agency(ies) is trying to influence through provision of goods and services, general regulation and financing ■ Outcomes are those events, occurrences, or conditions that are the intended or unintended results of government actions (OECD) Examples ■ reduction in maternal / child mortality ■ increase in export ■ decrease of unemployment ■ reduced crime levels ■ improved literacy rate of the population ■ reduction in fatalities from road accidents ■ land productivity improvement Health Examples (contd.) Life expectancy at birth/healthy life expectancy Incidence of all cancer/skin cancer Infant mortality rate Expenditure on prevention and health promotion Education & training Public expenditure on education as a % GDP Education levels Educational attainment of the adult population Adult literacy Work Unemployment rate Employment rate National income Examples (contd.) Real net national disposable income per capita Real GDP per capita Investment as % of GDP, by institutional sector Inflation rate Financial hardship Inequality of income distribution Productivity Labor productivity Expenditure on R&D The natural landscape Threatened species trend Fish stocks Forest resources Biodiversity Examples (contd.) Family community and social cohesion Suicide and drug-induced death rates Crime Violent and drug crime Killed and injured persons in road traffic Governance, democracy and citizenship Voter turnout and informal votes cast Women in Federal Parliament Cultural identity Attendance at cultural events State Measures and Change LATVIA Number of road accidents Number of injured Number of fatal injuries Number of registered cars 2004 2005 change % 48912 47353 -3.2 6416 5600 -12.7 516 442 -14.3 898145 966242 -7.6 State Measures and Change (contd.) ■ the same data can be called state measure or performance measure ■ it depends whether you attribute policy (outputs, regulation and finances) to it or not ■ if you do attribute – you are interested in change influenced by your policy (the question – whether different policy generates different result, e.g. marginal changes in result depending on inputs) ■+ - 0 State Measures and Change (contd.) LATVIA Number of road accidents Number of injured Number of fatal injuries Number of registered cars 2004 2005 change % 48912 47353 -3.2 6416 5600 -12.7 516 442 -14.3 898145 966242 7.6 ■ outputs ■ financing ■ regulation attribution of policy to change Different outcomes – results & impacts Impact or longer term effects (FO) Global objectives Results or direct & immediate effects (SO) Specific objectives Outputs Operational objectives Activities Inputs Results & impacts defined ■ Results relate to the direct and immediate effect brought about by a policy or program. They provide information on changes to, for example, the behaviour, capacity or performance of direct beneficiaries. Such indicators can be of a physical (reduction in journey times, number of successful trainees, number of roads accidents, etc.) or financial (leverage of private sector resources, decrease in transportation cost) nature ■ Impacts refer to the consequences of the program beyond the immediate effects on its direct beneficiaries. Two concepts of impact can be defined. Specific impacts are those effects occurring after a certain lapse of time but which are, nonetheless, directly linked to the action taken. Global impacts are longer-term effects affecting a wider population Examples Tax collection Economy Justice inputs Inputs Inputs Inputs outputs Tax audits Regulatory burden studies Prisoner rehabilitation results or SO Reduced tax evasion Better regulation Re-integration into community impacts or O Reduced shadow economy Enhanced competitiveness Reduced repeat crimes Results & impacts defined (contd.) ■ Results = short to medium term outcomes = specific outcomes ■ Impacts = medium to long term or final outcomes = overall outcomes ■ Sometimes immediate, intermediate and final outcomes Single outcomes and indexes ■ Sometimes indexes are used to describe and compare performance ■ Outcomes from several areas and sometimes even subareas are merged into one index ■ These are of particular use for policy planners ■ For example: the Enabling Trade Index Example: The Enabling Trade Index 2007/8 MARKET ACCESS Tariff & non-tariff barriers Proclivity or openness to trade BOARDER ADMIN. T&C INFRASTR. Efficiency of customs administration Availability & quality of transport infrastructure Efficiency of import – export procedures Availability & use of ICTs Transparency of boarder administration BUSINESS ENVIRONM. Regulatory environment Physical security Example: The Enabling Trade Index 2007/8 (contd.) Finland 7 / 5.61 Netherlands 11/ 5.51 United States 14 / 5.42 France & Ireland 19 / 5.20 Estonia 25 / 4.89 Lithuania 35 / 4.63 Turkey 38 / 4.53 Latvia 43 / 4.45 Poland 45 / 4.35 Romania 57 / 4.04 Moldova 62 / 3.88 Ukraine 68 / 3.77 Macedonia 81 / 3.58 BiH 89 / 3.47 Kyrgyzstan 109 / 3.03 Example: Tourism Competitiveness Index 2008 REGULATORY FRAMEWORK BUSINESS ENVIRON. & INFRASTR. Policy & regulations Air transport infrastructure Environmental regulations Ground transport infrastructure Safety & security Tourism infrastructure Health and hygiene Prioritization of T&T Price competitiveness HUMAN, CULTURAL & NATURAL RESOURCES Human resources National tourism perception Natural & cultural resources Example: The Tourism Competitiveness Index 2008(contd.) United States 5 33 1 12 France 12 13 5 28 Finland 16 7 18 33 Netherlands 19 22 15 25 Ireland 27 14 26 46 Estonia 28 32 25 34 Lithuania 51 57 43 61 Turkey 52 53 63 48 Latvia 53 60 41 77 Poland 63 63 62 60 Georgia 66 55 98 31 Romania 76 87 74 71 Ukraine 78 76 73 89 Macedonia 83 114 82 44 Moldova 95 99 100 83 Kyrgyzstan 102 111 104 84 BiH 104 101 94 108 Problems with outcome measurement ■ Timing ■ Data availability for measurement and transaction costs ■ Outcome (state) measures remain important even in the absence of policies and programs directed at them ■ Impact of other influences ■ Attribution Other influences Influence of other factors Degree of outcome attribution to outputs Attribution to outputs Attribution Immediate – intermediate - final Attribution modelling Outcomes IMMEDIATE Knowledge & understanding INTER MEDIATE Behavioural changes FINAL Changes in environment Attribution modelling (cont.) Outputs Inputs Outputs Target group Outcomes Short term Improved knowledge about impact of farming on environment Educational events Staff Finances Research Materials Control system Polluting farmers Improved understanding about the use of pollutants Medium term Systematic pollution monitoring in farms Long term Less polluting farming etc. Site visits / inspections Improved understanding about the ways to control pollution Changes to farming practices Improved environment Summarizing about outcomes ■ Outcomes are changes in the economic, physical, social and cultural environments which the state agency(ies) is trying to influence ■ Focus on change (+/-/0) in outcomes is what makes attribution between outputs and outcomes) ■ Outcomes can be of immediate, medium term and long term nature. More immediate they are – stronger the attribution to outputs ■ Outcomes can range from single to more compplex ones based on some kind of a model ■ Outcomes are particularly useful for policy planning and evaluation Contents ■ Outputs & inputs ■ Outcomes ■ Indicators ■ Logic model ■ Budgeting & PI ■ Future Indicators - definition ■ An indicator can be defined as the measurement of an objective to be met, a resource mobilized, an effect obtained, a gauge of quality or a context variable ■ Indicators demonstrate how to measure elements of program that you are interested in or program in general Different levels of indicators Single indicators Indicators on input What goes into the system? Which resources are used? Indicators on output Which products and services are delivered? What is the quality of these products and services? Indicators on intermediate outcomes What are the direct consequences of the output? Indicators on final outcomes What are the outcomes achieved that are significantly attributable to the output? Indicators on the environment What are the contextual factors that influence the output? Examples Description Indicators Output •Construction of a road Implementation: •financial: cost, state of progress •physical: km constructed, level of progress Result •Reduced journey time and •Accessibility (ESS)1 transport costs •Time savings (in min) •Cost savings (%) Specific impact •Increased safety •Increased flows of persons and goods •Traffic flows Global impact •Increase in socioeconomic activity •diversification of production •net job creation •Increased regional GDP per capita •and per occupied person. Important aspects of use of indicators ■ Output indicators ■ State indicators and baselines ■ Targets ■ Single and ratio indicators Output indicators ■ Volume or quantity of provision is the most often used and most relevant to budgeting. However, other output dimensions can have indicators, too: ■ Timeliness in provision ■ Quality of provision: satisfaction and/or comparison ■ Coverage, meting of demand and accessibility ■ Equity Example: technical certification of cars service ■ Volume: X number of cars per years ■ Coverage & demand: 100% every day ■ Timeliness in provision: less than ½ hour waiting, service within 1 hour ■ Quality: satisfaction of car users; quality in weakest areas improved; online registration etc. ■ Equity: special service for disabled State indicators and baselines ■ State indicators describe the state of affairs in a given area of policy concern – unemployment, GDP, mortality, lack of labour mobility, demand for a service exceeding supply etc. ■ Baseline indicators refer to the initial value against which inputs, outputs or outcomes are subsequently measured. ■ State indicator = photograph of situation; Baseline indicator = starting point of measuring impact of policy Targets ■ Show desired level of performance ■ Make sense only if time bound and have baseline ■ Customer satisfaction from a service increased from 25 to 75% by June 2009 ■ Every input, outputs and outcome can have indicators, but should it? ■ UK PSAs: from 300+ targets in 1998 to 30 targets in 2007 ■ Often, if you want to have loose connectionto accountability and control, simple expected performance projections can work better than targets Static and dynamic targets STATIC CONCEPT DYNAMIC CONCEPT Value Needs Value Target Needs Target Baseline Baseline Time Single and ratio indicators Ratio indicators Efficiency Costs/Output Productivity Output/Input Effectiveness Outcome/Objective (intermediate or final) Cost-effectiveness Input/Outcome (intermediate or final) These measures are valid only to the extent that there is a clear causal relationship Single and ratio indicators (cont.) Needs Objective Input Effectiveness Productivity Costs Cost-effectiveness Efficiency Output Outcomes Summarizing about indicators ■ Indicators are measures of change ■ Baseline and target data is important for good use of indicators ■ But be careful – not too many targets. When targets are set, make sure gaming potential is minimized Contents ■ Outputs & inputs ■ Outcomes ■ Indicators ■ Logic model ■ Budgeting & PI ■ Future Problems and needs Other influences Policy input Structure, institutional and managerial arrangements Objectives Program level Policy level Model for structuring performance information input Activities and processes Outputs Short to medium term outcomes Final outcomes Context Value to the nation Challenges Measurement challenges Measurement costs Relative magnitude Time delays Linkages to our actions Outcomes OUTPUTS IMMEDIATE INTER MEDIATE FINAL Bottom-up vs. top-down approach in modelling and planning ■ Outputs to outcomes (bottom-up) approach is most common to budgeting OUTPUTS Outcomes INTER MEDIATE IMMEDIATE FINAL ■ While outcomes to outputs (top-down) – in policy Policy objective Options Instruments ■ However, several OECD countries use the later in budgeting, focusing on outcomes and leaving connection to outputs loose Contents ■ Outputs & inputs ■ Outcomes ■ Indicators ■ Logic model ■ Budgeting & PI ■ Future Budgeting and performance information ■ Performance budgeting has been THE THEME for several decades now ■ It has not succeeded and no one does it right ■ There are two most discussed topics in this theme: ■ Outputs or outcomes focus? ■ Can everyone do it? Challenges ■ Informational challenges, e.g. inserting output and outcome information into budget documentation ■ Decision making challenges, e.g. how much difference PI makes Informational challenges ■ Information overload ■ Difficulties to delegate ■ Cost of data ■ Information and knowledge gaps ■ Principal agent problems Decisional challenges ■ Better information = better decisions – only assumption ■ Hard to attribute action to outcome change. Requires high quality policy analysis ■ Fixed budgets are not performance budgets ■ But non-fixed budgets can lead to gaming ■ Information asymmetry and principal agent problems ■ Time delays Outputs or outcomes? outcome s outputs Program budgeting Strategic planning Evaluation and spending reviews Attribution Need for strong policy analysis Outputs or outcomes? (cont.) ■ Performance budgeting is about using the budget to promote performance rather than a specific technique such as program or output budgeting. What all attempts at performance budgeting have in common is the attempt to increase the influence of policy analysis, strategic decision support, performance information in the budget system (Graham Scott) ■ Appropriate use of performance information by key actors at each stage of the budget cycle to inform their decisions concerning resource allocation and to improve efficiency of resources (World Bank) Can everyone do it? ■ Get the basic parts of the budget to perform first ■ Willingness and ability to execute voted money, to control expenditure ■ Sound financial accountability, respect for financial management procedures, and effective internal controls ■ Functioning government ■ Implanting of performance information in the budget process requires conductive environment – values, minimal degree of informality, less corruption, leadership etc. Where to start? 1. Aggregate Fiscal Discipline How much money you can spend or should save for the whole of Government? Macroeconomic policy as in Budget Memorandum 2. Allocative Efficiency 2.1 Inter-sectoral 2.2 Intra-sectoral In what areas we want / need to spend the money on – health, education, transport? Strategic policy documents of the Government In each area, what are the needs and priorities we want to spend money on – preventive health care, treatment etc.? Sector level policy documents 3. Operational Efficiency How well we spend the money we got? Could we spend it better? Monitoring and analysis Where to start? (cont.) Aggregate fiscal discipline However, any failure to successfully achieve macro fiscal goals will seriously undermine allocative and operational efficiency goals Allocative efficiency Operational efficiency Importance of PI What preconditions? (cont.) ■ Baseline projections ■ Trend analysis ■ Attribution to change What preconditions? ■ Get the basics right (execution, control, accounting, reporting) ■ Sort out budget process (actors, information flows, incentives and capacity) – little sense to attempt performance budgeting within environment where short term political priorities and deal making prevail over rational analysis of spending options ■ Focus on performance, and not performance information ■ Focus on process, not only targets. Process is as important as targets ■ Role and capacity of Ministries of Finance and centres of Government ■ Holistic rather than patchy approach Contents ■ Outputs & inputs ■ Outcomes ■ Indicators ■ Logic model ■ Budgeting & PI ■ Future Future - open questions ■ How does your PI dimension affect your budget process in terms of actors, information flows between them, capacity and incentives? ■ How to improve usage of PI in budgetary decision making? ■ How to improve measurement? ■ How to improve quality / user friendliness / usability of information? ■ How to get politicians to use PI in decision making? www.cpmconsulting.eu