National Indicators

Chapter profile

Table (1): Chapter profile and remit boundaries.

This chapter is not another numerical annex repeating the economy, health and energy chapters. Its role is to establish the “rules of the game” that preserve accurate definitions and comparisons when the Vision moves from sectoral chapters to a single implementation dashboard. If unemployment has a survey definition and formal employment an administrative source, they must not be merged into one figure. If water has separate indicators for supply, quality and interruption, increased coverage does not mean that water has become safe and continuously available.

Governing rule: a national indicator describes a specific phenomenon in a defined year and population or territory; it is not equivalent to an announcement, appropriation value or activity count.

1. Executive summary

A state can fulfil announced expenditure while deep development gaps persist. Iraq Vision 2045 therefore makes measurement a governance function, not merely a reporting function: reliable knowledge of conditions, a testable goal, a body able to address it, and publicly disclosed updates when circumstances change. The Vision’s aim is not the highest ranking on every global index. It is for Iraqis to experience sustained improvements in security, job opportunities, income, learning, health, water, electricity and state services, with those changes documented and attributed to their proper year and rationale.¹

Iraq already has a statistical foundation that must not be ignored. The 2024 population census established a population of 46,118,793, while the 2021 Labour Force Survey reported unemployment of 16.5% and a youth not-in-education-employment-or-training rate of 36.7%. Statistics for large industrial establishments in 2024 identified 916 operating establishments, and International Telecommunication Union data show internet use by 81.5% of individuals in 2024. These figures must not be combined into a false annual dashboard: each has a different year, coverage, definition and purpose.²

The Ministry of Planning has already worked on indicators and outcome matrices for the 2024–2028 National Development Plan and continued developing IDMS for project-data management in 2026. This Vision therefore proposes no new “bureau of numbers” outside the statistical system. It proposes a national reference register of indicators, managed statistically with sectoral bodies and connected to programmes and projects without conflating them.³

The chapter proposes 40 carefully defined diagnostic indicators in five areas, including 15 initial headline indicators for the concise national dashboard. This is an editorially selected set open to refinement after the 2027 audit, not a statement that the state has formally adopted it. Targets are distinguished as documented actual values, published government targets for specified periods, proposed policy targets, and indicators whose targets cannot be fixed before a baseline exists. This distinction prevents the Vision’s numbers from becoming claims about the present or unsupported promises about 2045.

Figure (2): Measurement is a management tool, not a publicity step at the end of implementation.

2. Why measure?

First, measurement distinguishes “the problem is large” from “how large, where, and for whom?” Second, it helps allocate scarce resources: would maintaining an existing water network deliver greater benefits than building a new plant? Does road improvement reach areas with high traffic fatalities? Third, measurement identifies an intervention’s impact and distinguishes coincident change from its cause. Finally, it enables institutions to learn from mistakes and limits narratives of unverifiable achievement.

Measurement can be harmful if it rewards behaviour aimed at “success in the number” rather than success in the function. Transaction times may fall because an office refuses difficult applications; school pass rates may rise while reading remains weak; exports of a hydrocarbon product may grow and be proclaimed diversification that never occurred. Outcome indicators therefore need accompanying measures of measurement integrity and quality that prevent costs being shifted to citizens or the environment.

3. The difference between a slogan, an indicator and an outcome

Table (2): Turning slogans into outcome indicators and tests against manipulation.

Every goal needs precise wording beyond a number. It must establish the desired direction, unit, denominator, period, geographical and social breakdown, producing body, evidence source and review date. A target is not a fact about the future: “raising reading proficiency to a specified level by a given year” is a conditional policy decision, whereas “proficiency in 2024” is a fact only if derived from a test with a published methodology.

4. National outcome architecture and the denominator principle

The Vision uses the chain “input → activity → output → outcome → impact”. Money and staff are inputs; training courses or network construction are activities; trained teachers or completed connections are outputs; changed student skills or service continuity are outcomes; changes in human productivity, health and life are impacts. These layers must not be combined into one indicator, nor should an easier indicator substitute for impact because impact is harder to measure.

Figure (3): The measurement chain; an output is not synonymous with an outcome.

Table (3): A practical example preventing confusion between expenditure and impact.

5. Rules for sources and baselines

Data from the statistical authority, the regional statistical authority, ministries and oversight bodies are preferred for matters within their mandates and according to their published definitions. International institutional data are used where comparison is needed or a consistent methodology is available but not published locally. Where series differ, no arbitrary average is taken; differences in coverage, year, denominator, prices, classification and revision year must be investigated. The source of an estimate is not necessarily the source of a census or administrative register.⁴

The 2027 national baseline is not a value from one year chosen for every indicator. It is a “set of multiyear baselines” retaining the latest valid measurement for each item. Survey data from 2021, such as NEET, may be used if their age is clearly stated; they must not be called Iraq’s situation in 2026. Unmeasurable indicators are labelled “unavailable / to be established in 2027–2029”, not assigned zero, because zero is a quantitative claim distinct from missing information.

Table (4): Indicator reliability levels and rules for using each category.

6. Iraq’s available baseline for 2021–2026

The aim is to provide selected reference points for designing the measurement system, not to produce another comprehensive development report. Values, years and methodological limits are presented. An annual series must not be inferred between two different measurements merely because two numbers exist, nor should a 2026 estimate replace completed national accounts for 2024 or subsequently published results of the 2024 census.

Table (5): Examples of documented baselines, not a complete national dashboard.⁵

These examples confirm that Iraq produces sound data in specific fields, but they are insufficient to evaluate the Vision as a whole. Operational indicators for time, quality, service continuity, the judiciary and projects are not available in a unified national form in the preceding chapters’ files. The first programme therefore builds measurement tools for them rather than substituting broad international rankings. The 2023–2024 multidimensional poverty survey should also be added as a perspective distinct from monetary poverty, not an arithmetic substitute.⁶

7. The national indicator hierarchy and naming system

The chapter proposes five stable families: G for security, sovereignty and governance; E for economy and production; H for human development; T for science and technology; and I for infrastructure, energy and services. Each card carries a code such as V2045-G01 and a definition version that never changes silently. A “●” marks 15 preliminary headline indicators, three in each family; the others support explanation and diagnosis. The total of 40 is a “foundational register”, not a final limit for every chapter and annex of the Vision.

The top-level dashboard publishes only 15 indicators so it remains readable and allows trade-offs to be explained. The second level publishes all forty with analysis and distribution, while annex V4-A01 retains broader sectoral cards, trends for 2030–2045, data-owner names and definition-change dates. The priorities among the 15 may be revised if experience in the early years identifies a better indicator of the outcome.

8. Security, sovereignty and governance indicators

Security and governance are measured not only by incidents and decisions, but by the legitimacy of rules, the state’s capacity to enforce them, deliver services and protect the public sphere, and the integrity of contracting and judicial processes. The register therefore combines direct harm indicators with measures of trust, enforcement and readiness. Security-readiness percentages revealing operational details must not be published, but aggregate outcomes and lawful oversight procedures should be.⁷

Table: Initial indicator register for domain G; ● denotes an initial headline measure.

The first phase defines serious-harm offences, response events and the judgment-enforcement register. Police arrival time is not the time taken to pass a report through a call centre; increased enforcement of judgments must not sacrifice defence rights or select only easy cases. Violence rates must be interpreted in light of reporting quality, because more reports may reflect growing trust rather than more crime.

9. Economic and production indicators

The national economy needs indicators that distinguish rising expenditure and prices from greater productivity, export capability and productive employment. Key measures concern non-oil public finances, real productivity and documented export diversification: a falling oil share of GDP caused by lower oil prices is insufficient evidence of successful diversification. Crude oil, petroleum products, other hydrocarbons, service exports and re-exports must be separated.⁸

Table: Initial indicator register for domain E; ● denotes an initial headline measure.

Real output is calculated at prices from a clearly specified base year, and series are revised if the national accounts base changes. Formal jobs are not simply one-day worker registrations, and corporate credit is not financing for productive investment without classification by purpose and maturity. E01 must be coordinated with V2-D04-C01 and V2-D04-C07 to avoid creating a competing non-oil revenue figure.

10. Human development and social justice indicators

Successful human development is determined not just by school, bed or training-course counts, but by people’s capacity to learn, work, enjoy health and live with dignity. This domain connects monetary poverty, educational capability, the transition from study to work, women’s participation and protection of vulnerable people. Each indicator is disaggregated by sex, age, governorate, urban/rural location and income where data quality and privacy permit.⁹

Table: Initial indicator register for domain H; ● denotes an initial headline measure.

Unemployment can improve when people leave the labour force; it must therefore be read alongside participation, NEET and formal employment. Likewise, lower monetary poverty does not eliminate shortages of water, education or health, so the measurement system retains multidimensional poverty separately. HCI describes potential productivity over a long horizon; it is not suitable for judging an education minister monthly.

11. Science and technology indicators

Technological transformation is not equivalent to buying devices, creating platforms or opening centres. Measurement must cover research expenditure classifiable under Frascati, innovation applied in firms, the effect of digital services on citizens, and resilience to incidents. Greater internet use is necessary but can coexist with major disparities in speed, quality and security. T03 therefore appears alongside I07, not in its place.¹⁰

Table: Initial indicator register for domain T; ● denotes an initial headline measure.

Digital-security measurement draws on the existing cybersecurity framework, but an international maturity score such as the Global Cybersecurity Index 2024 is not equivalent to a protection rate or recovery time. A single ranking must not replace a government-service measure or patent output. Sensitive internal operational indicators must be separated from aggregate public versions that do not expose response details or critical systems.

12. Infrastructure, energy and service indicators

Infrastructure and services are measured at the point of use, not at opening ceremonies. Handover of a power plant does not mean full supply; laying a pipe does not ensure safe water; announcing a highway does not prove lower journey costs; and upgrading irrigation equipment does not ensure water savings if the irrigated area expands without limits. Infrastructure indicators therefore require operating data, quality monitoring and household and enterprise surveys.¹¹

Table: Initial indicator register for domain I; ● denotes an initial headline measure.

Energy, water and transport data should be disclosed by governorate, corridor and beneficiary type whenever legally safe. Network sewerage coverage must be distinguished from actual wastewater treatment, and internet-user shares from guaranteed broadband service. The development plan’s irrigation-efficiency value is used only within its definition and 2028 horizon; it must not be extrapolated linearly to 2045.

13. Cross-sectoral indicators and risks

Outcomes across parts interact; they cannot be mechanically added. Water affects crops, electricity, health and cities; oil volatility affects budgets, exchange rates and investment; skills affect institutional performance, manufacturing and technology. Each cross-sectoral issue should therefore have a “risk indicator” and an “outcome indicator” with a coordinating body. It must not become a vague shared responsibility without an owner.

Table (11): Cross-domain indicators and the risks of superficial comparisons.¹²

The adopted method rejects a “Great Iraq composite score” in the foundational phase, because weighting security against health, water or productivity is a normative political decision, not an automatic statistical output. Five separate dashboards and consistent trends can be presented. If an academic composite index is later used, its weights, sensitivity, missing data and methodological changes must be published, while the original measures remain visible.

14. How are indicators calculated and interpreted?

Each indicator needs a formula and measurement point that enable reproduction. Its denominator must represent the exposed population or actual service, not make the result look better. Service time runs from a complete application, under a published definition, to the final decision; it must not begin after waiting that occurs outside the system. Completion rates are divided by all cases due, not only closed cases, with exceptions stated.

Table (12): Formulas, denominators and risks of misinterpretation.¹³

The number of observations, missing values and the source for age, household or project information should accompany the series, with periods of non-comparability stated. If “complete service” or the labour-survey method changes, retain the old version and publish a revised series or an explicit break whose ends are not forcibly joined. International comparisons also require clarity on whether the data are national observations or modelled estimates prepared by an international agency.

15. Geographical, social and temporal disaggregation

“Iraq as an average” may be insufficient for fair decisions. If water reaches 90% of households on average, for example—a hypothetical figure, not an Iraqi statistic—that does not establish access in a remote village or continuity during summer. The register therefore adopts minimum disaggregation by governorates within survey coverage, urban/rural location, sex, age, income or poverty, and disability where source methodology and sample size permit.¹⁴

Governorates and regions must not become vehicles for unfair comparison. A major city faces greater demand, while a peripheral area faces different geography, density and costs. Absolute figures, population rates, costs and service types should therefore be shown together. The region and areas excluded from some industrial, traffic and water reports must also be accounted for. An unmeasured cell is not zero and must not be filled with the average of neighbouring governorates.

Table (13): Essential disaggregation with confidentiality and sample-size safeguards.

16. Interim targets for 2030, 2035, 2040 and 2045

The Vision follows four decision points—2030, 2035, 2040 and 2045—as established in V4-D10-C01. But not every indicator can follow a linear numerical path. Some are best near zero, such as deaths and losses; some are measured every three or five years, such as learning and poverty; some depend on sample quality or population growth; and others take years to respond to reform. Each category therefore needs a “target methodology”, not one number imposed on all.

Table (14): Indicator trajectories across phases—directions, not invented figures for 2045.

Detailed targets for the 40 indicators should follow the first baseline-year audit or adoption of documented targets from specialist plans. In this edition, the Vision does not claim to know actual poverty, unemployment or electricity levels in 2045 and does not set fictitious benchmarks for indicators with undefined denominators. This caution protects the programme’s accountability: a sound target requires a known starting point, an implementation mechanism and a financing framework, not merely bold language in an announcement.

17. What we draw from the 2024–2028 plan’s targets

The 2024–2028 National Development Plan contains official short-term targets and estimates that must be presented as such, neither treated as realised facts nor extrapolated linearly to 2045. Announced targets include reducing unemployment by at least 30% from the plan’s reference levels, lowering poverty to no more than 15%, raising agriculture’s share from 2.4% to 2.9% and manufacturing’s from 1.8% to 2.2% under the plan’s definitions, and an expected private-investment contribution of approximately 34.8% of total investment during the plan period.¹⁵

Table (15): National Development Plan targets for 2024–2028 and how to use them without exaggeration.¹⁶

If actual results and a final government evaluation are issued in 2027 or 2028, they must enter the register as facts with their own source years, not as an automatic extension of 2024 plans. Differences in the definition of “industry’s contribution” across series and national accounts must be reconciled before claiming that a target was met. Lower unemployment may also coincide with lower participation, so NEET, participation and formal private employment remain mandatory accompanying indicators.

18. How do we avoid improving the indicator while losing the outcome?

There are three major measurement risks. First is Goodhart’s law: when a measure becomes the sole administrative target, an institution may change to produce the number rather than the service. Second is causal attribution error: household income may rise because oil prices increased, not because of a training programme. Third is distributional error: the national average improves while the gap between two governorates widens. No programme should therefore receive a final success judgment from one number detached from a countercheck.

Table (16): A countercheck accompanying every headline indicator.

Professional staff must be protected from incentives to hide difficult cases or alter numbers; people reporting data errors need protection and access to independent correction. Corruption cannot be inferred from every percentage change: a census update or newly discovered classification error may explain it. But the institution must provide a public revision log explaining what changed, when, why and how it affects comparisons across years.

19. A metadata card for every indicator

Every indicator requires a mandatory metadata card with at least 17 fields. The card is stored in the central register under a stable code and dated versions, linked to the data portal and annual national report. No value may appear on the Council of Ministers’ dashboard if its source, denominator, measurement year or certifying body is unknown.

Table (17): Minimum fields for a national indicator card.

Applied example H01: scope, ages 15–24; numerator, young people not in work, education or training during the reference period; denominator, all surveyed members of that age group; source, the Labour Force Survey; latest documented baseline here, 36.7% for 2021; updating through a new survey; disaggregation by sex and location; and deviations referred to education, training and labour bodies according to cause. “Annual increases” between 2021 and a year without a survey must not be measured by presenting projections as realised outcomes.¹⁷

20. Statistical governance and institutional ownership

Existing institutions produce the data: the national and regional statistical authorities within their mandates; the Ministry of Planning for national plans; Finance for expenditure; the central bank for monetary data; ministries and service facilities for operational data; and judicial and security bodies for information that may be shared. Iraq Vision 2045 does not give a political office or minister sole ownership of a statistical definition. Changes require a documented request, methodological review and independent technical acceptance.¹⁸

Figure (4): Separating data production from validation and from accountability of implementers.

Indicators requiring cooperation between two bodies need a protocol identifying who owns the primary register, who publishes the national value, who must correct the service, and who preserves version history. If electricity carries a household-supply indicator, the Ministry of Electricity alone cannot adequately audit household satisfaction and interruptions. The statistical body can conduct a verification survey and analyse geographical bottlenecks. V4-D10-C03 translates this relationship into administrative performance agreements within the law; this chapter does not repeat that functional design.

21. Data quality and independent audit

Six dimensions of data quality are adopted: fitness for purpose, measurement accuracy, coverage completeness, consistency and comparability, publication timing, and verifiability and interpretability. Each needs a test: does the indicator reflect citizens’ lives? Are all groups covered? Has the denominator changed? Is publication too late to inform decisions? Can an independent reviewer reproduce the calculation from a secure database or aggregate table?¹⁹

Table (18): The technical and legal verification gate before publication.

For indicators generated by daily administration, independent review of a transaction sample is more useful than checking the table’s appearance alone. Reviewing 100 random transactions, for example, is not the objective itself: a sample must be designed for service volume, error risk and the required confidence level. The size of 100 is illustrative, not a national legal or statistical standard. After review, the body publishes the error rate, scope of correction and date of series reproduction.

22. The national dashboard and its three levels

The dashboard has three layers: a concise public page for citizens containing the fifteen headline indicators, their directions and source years; a sectoral page containing all forty, their disaggregation, data quality and explanations of deviations; and a detailed government operational layer whose sensitive information is not published and which supports intervention in projects and services. These layers serve three different functions without producing three competing definitions.²⁰

Table (19): Three presentation layers from one definition.

Red/amber/green signals must not appear without dates or criteria. “Red” may mean delay against a documented target, a data-quality problem or a risk to the legality of a decision; these are different conditions. The dashboard should therefore separate performance colour from source-quality colour and show the owning body, corrective action and due date, with room for explanation that does not become empty rhetoric.

23. The updating, correction and reporting cycle

The indicator cycle needs known timing: monthly operational monitoring of services and projects; quarterly risk and dependency updates; annual or survey-based publication of poverty, employment, learning and health data as sources permit; and independent evaluation after each phase. A national report is not correct simply because it was issued on time. If an essential measurement is missing, the status must be declared unknown rather than assigned undocumented success.

Figure (5): The evidence-based management cycle does not end with uploading a report.

Every review documents three possible causes of change: an established government intervention, external factors such as oil prices and rainfall, or a methodological change in the data. Not every improvement can be attributed to a Vision programme merely because it occurred during a government’s term. To estimate a particular programme’s impact, before-and-after measurement, comparison groups where feasible, or an ethically and methodologically legitimate quasi-experimental design are recommended. Limits of certainty should be published rather than declaring unprovable causation.

24. Implementation programmes for measuring the Vision

This architecture requires seven limited operating programmes, not a new ministry. Each starts with a tangible product during 2027 and may be reduced or merged into an existing programme if a mandate and cost study establishes that its function is already performed. The programmes primarily address missing data, institutional connections, analysis and publication of evidence, rather than purchasing platforms parallel to existing systems.

M10-01 | National reference register of codes and definitions

Task: a dictionary of 40 cards, their versions and a 2027 change log. Implementation begins within existing bodies’ legal mandates, led by Planning/statistical bodies; no new organisation should be created before comparing the costs of improving the existing system and replacing it. The first success indicator is a published, reproducible data product or audit report, not counts of courses and workshops.

M10-02 | Baseline reconciliation

Task: reconcile sources, years and denominators and disclose differences. Implementation begins within existing bodies’ legal mandates, led by the statistical authority and competent bodies; no new organisation should be created before comparing the costs of improving the existing system and replacing it. The first success indicator is a published, reproducible data product or audit report, not counts of courses and workshops.

M10-03 | Unpublished service measures

Task: pilots measuring time and quality, with independent cross-checking for every critical service. Implementation begins within existing bodies’ legal mandates, led by service bodies/Planning; no new organisation should be created before comparing the costs of improving the existing system and replacing it. The first success indicator is a published, reproducible data product or audit report, not counts of courses and workshops.

M10-04 | Connected national indicator portal

Task: a public dashboard with source links and safe downloadable tables. Implementation begins within existing bodies’ legal mandates, led by Planning/competent digital-transformation bodies; no new organisation should be created before comparing the costs of improving the existing system and replacing it. The first success indicator is a published, reproducible data product or audit report, not counts of courses and workshops.

M10-05 | Surveys of social and human-development gaps

Task: regular surveys of learning, youth, poverty and service experience. Implementation begins within existing bodies’ legal mandates, led by statistics/health/education/labour; no new organisation should be created before comparing the costs of improving the existing system and replacing it. The first success indicator is a published, reproducible data product or audit report, not counts of courses and workshops.

M10-06 | Indicator quality audit and protection

Task: a sample-audit protocol, version register and publicly disclosed corrections. Implementation begins within existing bodies’ legal mandates, led by statistical bodies and the audit board according to their powers; no new organisation should be created before comparing the costs of improving the existing system and replacing it. The first success indicator is a published, reproducible data product or audit report, not counts of courses and workshops.

M10-07 | Evaluation, learning and disclosure

Task: outcome and evaluation reports at the 2030, 2035, 2040 and 2045 milestones. Implementation begins within existing bodies’ legal mandates, led by Planning/the centre of government/the evaluator; no new organisation should be created before comparing the costs of improving the existing system and replacing it. The first success indicator is a published, reproducible data product or audit report, not counts of courses and workshops.

25. A matrix of responsibilities, financing and duration

The table clarifies responsibility without claiming that a government body has accepted an assignment. The “lead body” is a policy proposal subject to legal mandate, government approval and resource assessment. The cost categories below are relative guidance, not dinar estimates or approved commitments. One-off investment must be separated from operations, personnel and recurring audit.

Table (20): Priorities, responsibility and qualitative financing matrix.

Development should initially be financed through existing systems, information, survey and programme functions within the budget and under the law. The technical shortfall justifying additional financing or a disclosed technical-support partnership is then identified. The Vision establishes no aggregate national cost without a study and does not treat international financing as a permanent substitute for the national statistical budget. An operating, maintenance and staffing allocation must precede platform development so that data do not become a project that stops after contracting.

26. Measurement risks and legal protection

The indicator system faces politicisation, selective classification, pressure on statisticians, excessive reliance on precise-looking but unfounded digital figures, conflicting source definitions, privacy violations, security-data leakage and the use of international rankings as the sole government standard. Addressing these risks requires neither blanket secrecy nor unrestricted publication, but equal, professional publication rules, a right to correct figures and lawful restrictions on sensitive data.²¹

Table (21): Risks to measurement-system quality and ethics.

27. What a successful measurement system looks like in 2045

Success in 2045 is not “a hundred indicators, all green”. Water or health problems may persist despite improved state resources. Success means knowing their extent through current, specific figures, acknowledging the cause of deviation, changing a programme when its hypothesis fails, and preserving series so researchers, citizens, the judiciary and parliament can review them within their respective mandates. The executive must stop treating changes in statistical figures as personal challenges.

The greater success is using data to eliminate false trade-offs: supporting water-efficient farmers rather than uncontrolled expansion; protecting maintenance of an operating water network rather than opening an unfunded plant; improving education quality in a low-performing neighbourhood rather than comparing building counts; increasing genuine export capability rather than shipments of a raw material. Statistics then become part of building national capability, not an archive of numbers.

Governing rule: in 2045, the Vision can be corrected because statistical facts do not change with the government, while recommendations remain reviewable because they are not statistical facts.

28. Conclusion and the bridge to the performance-based state

The preceding chapter established how transformation proceeds and the conditions for crossing its phases. This chapter establishes a shared language for measuring outcomes: an indicator, not a slogan; a source, not an impression; a denominator and year, not a floating percentage; a disclosed gap, not an invented value; and a corrective decision, not a perfunctory report. This language becomes 40 foundational indicators in five thematic families, including 15 proposed headline measures, with rules protecting comparison, distribution and privacy.

Chapter Three, V4-D10-C03, “The Performance-Based State”, takes up a new question: after defining what must be measured, who is legally responsible for changing it? How does an outcome affect a ministry’s plan, budget, accountability and annual report, and when does management warrant a reward, correction or scrutiny? The next chapter neither repeats statistical definitions nor invents new penalties; it builds on the register to establish a constitutional, professional outcome-management system.

29. The measurement-gap register

• Governance and the economy: establish baselines for security harm, the judiciary, service time, productivity and real exports under competent bodies’ supervision in 2027–2029.

• Human development and technology: new surveys of employment, learning, health, research, innovation and fully completed digital services, without filling gaps with estimates.

• Infrastructure and services: measure electricity interruptions, household water quality, sanitation, transport and irrigation, linking operating records with user surveys.

• Common condition: no distant numerical target before establishing the baseline year, denominator, source, data owner and verification method.