Availability: Budget and spending data

To what extent is government budget and spending information (budget execution) available as structured open data?

Definitions

Definitions and Identification

Most governments carry out an annual budget process, involving proposing and approving a budget, and reporting spending against that budget. The Open Budget Survey provides a regular assessment of how transparent this process is, with a focus on the documents involved in the budget process. This indicator complements the evidence collected by the Open Budget Survey by looking specifically at whether structured data is available on: proposed budget, amended budget, approved budget (the budget formally agreed upon by the appropriate legislative process in the country), government spending, extrabudgetary spending, social security spending, and public corporation spending.

Budget documents often present summary tables that describe the proposed and approved budgets. The budget part of this indicator asks whether the data behind such summary tables is available. This will usually take the form of a dataset organized with rows that contain the value of each budget line (possibly including proposed, agreed, and amended values), along with columns that provide classifications of the line.

Reporting on spending or budget performance should involve showing the total spend against each line of a budget. It may extend to showing information on how the goods, works, or services to be funded by that budget line have been delivered. Structured data on performance might include a dataset, or a column in a budget-related dataset, that shows how much has been spent to date against each budget line.

Further, this indicator assesses the presence of structured budget classification data consistent with internationally agreed standards. It asks for checks on four kinds of classification, though researchers are encouraged to add notes in the justification about other forms of notable classification used (e.g., geographical):

• Administrative classification identifies the entity that is responsible for managing the public funds described by a budget line, such as the ministry of education and health or, at a lower level, schools and hospitals.
• Economic classification identifies the type of expenditure incurred; for example, salaries, goods and services, transfers and interest payments, or capital spending.
• Functional classification categorizes expenditures according to the purposes and objectives for which they are intended.
• Program classification categorizes expenditures according to the programs used to enact public policies, thus aligning policies and programs with administrative structures.

(Sources: IMF Technical Note on Budget Classification and IMF's Fiscal Transparency Handbook (2018))

Data should be disaggregated both at the transaction level and with regard to cross-cutting programs. Transaction-level spending data records spending at a granular level, often with many rows of transaction data for each line of the budget; transactional data may include details of each counterparty (e.g., buyer and supplier).

Two sub-questions of this indicator ask researchers to assess whether the data has identifiers or other features that make it easy to connect:

• Budget and performance/spending
• Budget and procurement

This may take the form of clearly documented classifications that uniquely identify budget lines, or the presence of stable unique identifiers for budget lines.

Research Guidance

Starting points

• Sources:
  • The Open Budget Survey contains assessments by budget experts on the availability of budget documents. Country results pages contain full researcher responses for each question along with the URLs to data that Open Budget Survey researchers identified. Check carefully to validate the technical assessments made by OBS researchers. This will usually involve opening and examining linked files, and checking if there are other alternative sources of information if the data linked from the OBD survey presents only summary tables.
    • Question EB-5 asks about the availability of machine-readable data on enacted or approved budgets.
    • Question IYRs-5, MYRs-5 and YER-5 ask about the availability of machine-readable in-year, mid-year, and year-end reports that may contain data on budget performance.
    • Question GQ-1b asks about the presence of a consolidated dataset of budget information.
  • The World Bank produced a dataset (last updated 2017) with details of country's financial management information systems (FMIS), whether or not they have public data, and where it may be located. This can provide a starting point to identify current budget data sources.
  • The BOOST Public Expenditure Database contains details of World Bank–supported budget data publication for a number of low- and middle-income countries.
  • The Managing COVID Funds project of the International Budget Partnership, published May 2021, includes information about pandemic-related updates to budgets and spending (see particularly the answers to questions 1 and 2 in your country's questionnaire), as well as the guidance and information countries have made available for budget and spending related to emergency fiscal policy packages.
• Search:
  • "Open budget data" + [country]
• Consult:
  • Open data advocates.
  • Investigative journalists who report on government budgets and expenditures or public finance more broadly, particularly any who appear to be using aggregate public finance data.
  • Officials of civil society organizations that advocate for transparency and accountability regarding public finance.

What to look for?

To complete the assessment for this question you will need to access and explore the available data. This may involve running queries on datasets to check the variety of fields included. Look for well-structured data that could be accessed or read into a database row by row.

Look for evidence that can answer the following questions:

• What budget data is available? Does the country publish proposed, amended, and approved budgets, not just as summary tables, but as the data itself?
• What spending data is available? Does the country publish government spending data, both in gross terms and on an accrual basis, including spending in annual and multi-annual investment projects? Does it do the same for extrabudgetary spending, social security spending, and public corporations spending?
• What types of classification are used? Administrative, economic, functional, program? Are these consistent with internationally agreed standards?
• Is the data disaggregated both at transaction level and for cross-cutting programs?
• Does the data include common identifiers that support easy analysis across budget and performance, budget and project?

National and sub-national considerations

This indicator asks you to carry out your assessment for the national government, but where national datasets also include sub-national and local government spending, or allow data from these layers of government to be aggregated together following common standards, this can be indicated in the "How comprehensive is the data assessed?" sub-question for this indicator.

Question sub-components

Show/hide supporting questions

Existence

• Is this data available online in any form?
  • Data is not available online.

    Supporting questions: Are there other offline ways to access this data in the country? (e.g., attending an office to inspect it).

  • Data is available, but not as a result of government action.

    Supporting questions: If government is not providing access to data, how is this data available? Please provide a URL(s) for where this data can be found.

  • Data is available from government, or because of government actions.

    Supporting questions: Please provide a URL(s) for where this data can be found.

Elements

• Kinds of data:

• There is structured data available on the executive budget proposal in gross terms, including spending on annual and multi-annual investment projects. (No, Partially, Yes)

  Supporting questions (conditional)

  If Partially or Yes: If budget and spending data can be found in multiple datasets, please provide the URL(s) where data on the executive budget proposal is located.

  If Partially: Please briefly explain your 'Partially' answer.

• There is structured data available on amended budgets (when applicable) or amendments of the enacted budget. (No, Partially, Yes)

  Supporting questions (conditional)

  If Partially or Yes: If budget and spending data can be found in multiple datasets, please provide the URL(s) where data on amended budgets is located.

  If Partially: Please briefly explain your 'Partially' answer.

• There is structured data available on the approved or enacted budget in gross terms, including spending on annual and multi-annual investment projects. (No, Partially, Yes)

  Supporting questions (conditional)

  If Partially or Yes: If budget and spending data can be found in multiple datasets, please provide the URL(s) where data on the approved or enacted budget is located.

  If Partially: Please briefly explain your 'Partially' answer.

• There is structured data about government budget execution or spending, in gross terms and on an accrual basis, including spending in annual and multi-annual investment projects. (No, Partially, Yes)

  Supporting questions (conditional)

  If Partially or Yes: If budget and spending data can be found in multiple datasets, please provide the URL(s) where data on government spending is located.

  If Partially: Please briefly explain your 'Partially' answer.

• There is structured data about the government's extrabudgetary funds spending, in gross terms and on an accrual basis. (No, Partially, Yes)

  Supporting questions (conditional)

  If Partially or Yes: If budget and spending data can be found in multiple datasets, please provide the URL(s) where data on extrabudgetary spending is located.

  If Partially: Please briefly explain your 'Partially' answer.

• There is structured data about the government's social security spending, in gross terms and on an accrual basis. (No, Partially, Yes)

  Supporting questions (conditional)

  If Partially or Yes: If budget and spending data can be found in multiple datasets, please provide the URL(s) where data on social security spending is located.

  If Partially: Please briefly explain your 'Partially' answer.

• There is structured data about public corporations' spending, in gross terms and on an accrual basis. (No, Partially, Yes)

  Supporting questions (conditional)

  If Partially or Yes: If budget and spending data can be found in multiple datasets, please provide the URL(s) where data on public corporations' spending is located.

  If Partially: Please briefly explain your 'Partially' answer.

• Data fields and quality:

• Budget entries have administrative classifications to an internationally agreed standard. (No, Partially, Yes) Answer 'Partially' if there are administrative classifications, but their alignment with international standards cannot be confirmed.

  Supporting questions (conditional)

  If Partially or Yes: If budget and spending data can be found in multiple datasets, please provide the URL where administrative classifications are located.

• Budget entries have economic classifications to an internationally agreed standard. (No, Partially, Yes) Answer 'Partially' if there are economic classifications, but their alignment with international standards cannot be confirmed.

  Supporting questions (conditional)

  If Partially or Yes: If budget and spending data can be found in multiple datasets, please provide the URL where economic classifications are located.

• Budget entries have functional classifications to an internationally agreed standard. (No, Partially, Yes) Answer 'Partially' if there are functional classifications, but their alignment with international standards cannot be confirmed.

  Supporting questions (conditional)

  If Partially or Yes: If budget and spending data can be found in multiple datasets, please provide the URL where functional classifications are located.

• Budget entries have program classifications according to an internationally agreed standard. (No, Partially, Yes) Answer 'Partially' if there are program classifications, but their alignment with international standards cannot be confirmed.

  Supporting questions (conditional)

  If Partially or Yes: If budget and spending data can be found in multiple datasets, please provide the URL where program classifications are located.

• Information about individual financial transactions or expenditures is available at the most disaggregated level of the economic classification level. (No, Partially, Yes)

  Supporting questions (conditional)

  If Partially or Yes: Please, briefly explain your answer.

• Data is disaggregated by cross-cutting programs, or issues such as SDGs, climate action, gender budgeting, etc, (No, Partially, Yes)

  Supporting questions (conditional)

  If Partially or Yes: Please, briefly explain your answer.

• The data contains common identifiers to connect budget and budget performance data. (No, Partially, Yes)

  Supporting questions (conditional)

  If Partially or Yes: Please, briefly explain your answer.

• The data contains identifiers that can be used to connect budget data with data on major projects (e.g., infrastructure construction) and procurement processes. (No, Partially, Yes)

  Supporting questions (conditional)

  If Partially or Yes: If budget and spending data can be found in multiple datasets, please provide the URL where these identifiers are located, and briefly explain your answer.

• Data openness, timing, and structure:

• Dataset is available free of charge. (No, Partially, Yes)

  Supporting questions (conditional)

  If Partially: Please briefly explain your 'Partially' answer.

• Data is openly licensed. (No, Partially, Yes)

  Supporting questions (conditional)

  If No: If there are explicit restrictions placed on re-use of the dataset, briefly describe those here.

  If Partially or Yes: If the data is provided with an explicit open license, please provide the name of the license, or a link to it here.

• Data is available in all the country’s official or national languages. If the country has no official or national languages, data is available in the major languages of the country. (No, Partially, Yes) Assess this against the list of official, national, or in-use languages you provided as part of your response to the governance indicator that asks, "To what extent do relevant laws, regulations, policies, and guidance require that data collection and publication processes be available in the country’s official or national languages?"

  Supporting questions (conditional)

  If Partially or Yes: Please briefly describe the language coverage available.

• There are accessible and open official tools available to help users explore data. (No, Partially , Yes) Answer 'Partially' if tools make it possible to get at extracts of data without having to download a full dataset. Answer 'Yes' if there is an interactive tool that displays user-filtered extracts of the data to answer simple questions without downloading data at all.

  Supporting questions (conditional)

  If Partially or Yes: Please provide URL.

  If Partially : What are the main barriers to accessibility and usability?

• Data is timely and updated. (No, Partially, Yes)

  Supporting questions (conditional)

  If Partially or Yes: When was the most recent update to this dataset?

• Historical data is available that allows users to track change over time. (No, Partially, Yes)

  Supporting questions (conditional)

  If Partially: Please briefly explain your 'Partially' answer.

  If Partially or Yes: For what time period(s) (e.g., start and end dates) is data available?

• Data is provided in machine-readable format(s) (No, Partially, Yes)

  Supporting questions (conditional)

  If Partially or Yes: Please provide a URL where this machine-readable data can be found. (Additional URLs can be included in the justification and supporting evidence)

  If Partially or Yes: Please provide a comma separated list of the formats available? (E.g. csv, json)

  If Partially: What prevents you from assessing this data as fully machine-readable?

• The machine-readable dataset is available as a whole (No, Partially, Yes) Answer no if it's only possible to access individual records; Answer partially if it's possible to export extracts of the data; Answer yes if there are bulk downloads or APIs providing access to the whole dataset without financial, technical or legal barriers.

  Supporting questions (conditional)

  If Partially or Yes: Please provide a URL where bulk download access is available or described.

  If Partially or Yes: If bulk access is provided through an API, please provide a link to where the API is described.

  If Partially: Please briefly explain your 'Partially' answer.

• Negative scoring:

• This information is missing required data. (There is no evidence of data gaps., There is evidence that a portion of mandated data is missing., There is evidence of widespread omissions in mandated data.) In cases where the indicator itself identifies a dataset(s) to assess against or a separate governance indicator has asked you to determine data requirements of a relevant governing framework, assess against that. In cases where there is no such identified dataset(s) or related governance indicator, assess based on the parameters laid out in the publication of the information (e.g., are some fields entirely empty when they shouldn't be?), your local knowledge (e.g., if the data is supposed to include information for all public officials, does the number of total entries look right?), and any broader research you may have done for this theme (e.g., have media articles decried the incompleteness of the data?).

  Supporting questions (conditional)

  If There is evidence that a portion of mandated data is missing. or There is evidence of widespread omissions in mandated data.: Please briefly explain.

• The availability of this data has been affected by government response to COVID-19. (No, Partially, Yes)

  Supporting questions (conditional)

  If Partially or Yes: Please briefly describe how COVID-19 affected the availability of this data.

Extent

• How comprehensive is the data assessed?
  • The data assessed covers one or more national government agencies, but there are many agencies or local government units without data available.

    Supporting questions: Please, briefly explain your answer.

  • The data assessed covers one or more national government agencies and there is similar data available for most other parts of national government.

    Supporting questions: Please, briefly explain your answer.

  • The data assessed covers one or more national government agencies,, and there is similar data available for the majority of the public sector (including sub-national government, state-owned enterprises or corporations, and extrabudgetary funds).

    Supporting questions: Please, briefly explain your answer.

  • The data assessed covers the majority of the public sector (including sub-national government, state-owned enterprises or corporations, and extrabudgetary funds; i.e., a consolidated dataset).

    Supporting questions: Please, briefly explain your answer.

Indicator Justification

Public financing is critical to achieving the 2030 Agenda. As the International Budget Partnership explains, budgets offer a concrete means to track a country's commitments to achieving the goals, while information on spending reveals whether countries have followed through on these commitments (2017:1–2). Transparency in public finance supports delivery of SDG 16 on effective, accountable, and inclusive institutions. Indicator 16.6.1 investigates primary government expenditures as a proportion of original approved budget, by sector (or by budget codes or similar). Similarly, indicators for SDG 1 on No Poverty require detailed information on government spending and resource allocation, particularly those for targets 1.a and 1.b.

While transparency has long been an important principle of public financial management, increasingly fiscal transparency efforts have emphasized providing not only fiscal documents, but also disaggregated data. Perhaps the most impactful use of public finance data is the improvement of public financial management and budget allocation. Data can be used to support gender budget analysis, green budget analysis, and evaluation of the impact of fiscal policy on minorities and marginalized groups. This indicator thus examines the extent to which government budget and spending information—also known as budget execution—is available as structured open data.