Reading Official Statistics and Open Data in Türkiye: A Practical Guide for 2026

سفير طرابزون · Oct 8, 2026

Reading Official Statistics and Open Data in Türkiye: A Practical Guide for 2026

A public-affairs post may quote a correct figure yet give it a misleading headline by overlooking its period, location or definition. Two screenshots of tables may also differ because the statistics were revised, rather than because the data contradict each other. Good interpretation begins with the question you want to answer, followed by the figure’s definition, how it was produced and which version you are using, before drawing conclusions about a decision or public service.

This practical guide for 2026 explains how to make your interpretation open to scrutiny; it is not a list of current economic figures. Official pages were reviewed up to 8 October 2026, including the Turkish Statistical Institute’s (TÜİK) revision plan for 2026. The numerical examples are explicitly hypothetical, so that teaching examples are not mistaken for official findings about a particular city or institution. The aim is to help you examine each new table when it appears, rather than rely on a remembered figure from an earlier year.

Archive photograph of Ankara taken from the International Space Station on 11 April 2009; this general view of the capital does not depict its built environment in 2026.

Begin with a question the data can answer

First identify what you want to measure. “Has the service improved?” is a broad question. “Has the number of completed transactions changed over the same period?” or “Has the average waiting time fallen, using the authority’s definition?” is more directly measurable. A complete answer may require several indicators: more transactions do not necessarily mean shorter waits or greater user satisfaction.

Choose the appropriate geographical level and time period. If the discussion concerns a neighbourhood, do not present a national rate as a direct description of that neighbourhood. If it concerns a full year, do not use a single month without saying so. In your research notes, record the subject, place, period, population covered and unit. These few lines help you reject a figure that does not answer the question, even if it appears on a familiar official website.

Distinguish describing what happened from explaining why. A table may show that a value changed, but cannot alone prove that a particular decision caused the change. The definition, coverage, number of service users or other circumstances may have changed. When an explanation is needed, examine suitable additional information and present your conclusion as analysis, rather than as a statement issued by the institution that published the figure.

Where to find the release and table

Start at TÜİK’s official website, then choose the statistics or portal relevant to your subject. The site links to its data, geographical information, population and official statistics portals, among others. At the October 2026 review, the official data link led to veriportali.tuik.gov.tr. Following the institution’s own link reduces the risk of relying on an outdated interface or a copied page in search results.

The institute’s access guidance identifies several types of material, including statistical releases, tables, databases, reports and metadata. A release summary is insufficient if you need detail it does not contain, although it may be enough for a broad question with a clear unit. Open the accompanying table when comparing years or groups, then read its footnotes.

Other public institutions may publish statistics within their own remits. Record both the producer and the organisation displaying the data if they differ. A page bringing together indicators from several sources does not mean all its figures share one methodology. Find each indicator’s definition and source, and establish whether it is an official statistical series, an administrative record or a separate study before combining it with another indicator in a comparison.

Three dates that should remain distinct

A figure has a reference period describing what it measures, a publication date when it became available and, sometimes, a revision date when its value or presentation changed. The year you downloaded a file is not a substitute for the data year. In 2026 you might open a report covering 2025, or find a series spanning many years that has recently been revised. Where the distinction matters, state clearly which period the data cover and when they were published.

Imagine a release on the previous year’s transactions appearing early the following year. Its recent publication does not make it “this year’s number of transactions”. A page may also receive a technical update without every table on it changing. Check the table title, period column and update note rather than relying solely on a date in the website footer.

Keep a preliminary figure’s status when quoting it. A release may describe a value as an estimate, a provisional result or a forecast; these terms carry meaning. An observed count differs from a projection in a future plan. Retain the source’s own degree of certainty, then explain what can be compared and what requires a later release.

Metadata explain how to read the figure

The institute explains that metadata give data meaning and describe their definition, source, quality, production and availability. They may cover the statistical unit, geography, classifications, methodology and revisions. Read this material before relying on an apparently simple comparison, particularly where a short indicator title conceals conditions determining who is included and excluded.

For example, “number of applications” may count submitted forms rather than people, and one person may submit several applications. “Services delivered” may also refer to different units within the same institution. Do not resolve this ambiguity by assuming every title uses a word’s everyday meaning. Find the original unit and write it beside the figure in your notes, so that transactions do not become people in the final headline.

When comparing figures, establish whether they describe all recorded units or a survey based on a sample. A large table does not prove that everyone is covered. Retain any limitations, errors or collection methods specified by the source. A result can be highly useful for an appropriate question while remaining insufficient for a narrow breakdown it was not designed to measure.

Counts, percentages and percentage points

In a hypothetical example, an authority records 200 applications out of 1,000 in a particular category: a share of 20%. It then records 240 out of 1,200: the count rises, but the share remains 20%. Reporting only the increase in numbers could suggest that the category has become more prevalent, although its proportion is unchanged in this example. State the denominator used to calculate a percentage when it affects the interpretation.

In another hypothetical example, a rate rises from 20% to 25%. That is a difference of five percentage points, but a relative increase of 25% compared with the original value. Both calculations are correct, yet answer different questions. Choose the expression you mean; do not write “rose by five per cent” when you intend only the difference in percentage points.

Check the unit too: thousands or individual people, a monetary amount or an index, a mean or a median. A small number of large values may affect the mean, whereas the median describes a different feature of the distribution. If the source reports one, do not silently substitute the other. A sentence explaining the unit is preferable to an eye-catching headline that permits conflicting mathematical interpretations.

Archive photograph of the Turkish Constitutional Court building from October 2020; it does not document a judgment or hearing in 2026.

What does it mean if a series was revised in 2026?

TÜİK publishes a revision policy and a plan for the current year; its 2026 plan covers both routine and major revisions. Revisions may relate to changes in source data, classifications, definitions, coverage, methodology or base year, or to seasonal and calendar adjustment. Begin by identifying the type announced and how far back it extends through the series, rather than offering a general accusation or defence.

Keep the date of the version you used. If you replace an old value with an updated one in a comparison, check the other periods so that you do not mix two versions of the same series. Downloading the entire available series at once may help, rather than copying one year from an old publication and another from a new table. Record that the values come from the version available on a specific date.

Suppose you built a chart from a saved table, then notice different values when reopening it. Compare the footnote, revision policy and series name, and record both values and their dates. If the reason remains unclear, ask the institution where it is explained. A difference merits investigation but does not explain itself. Careful interpretation sets out the documented change and its limits before drawing conclusions about motives or effects.

Public access and microdata

A public table does not imply open access to the individual records behind it. The institute explains that microdata concern statistical units and are subject to confidentiality safeguards and access arrangements. Researchers face different procedures, categories and modes of access: a link is neither an unrestricted download for everyone nor a guarantee that an application will be accepted. If you need this level of detail, read the relevant section’s current project, eligibility and use requirements.

For a public discussion, an aggregate table may be more appropriate than seeking individual records. Consider what extra detail would contribute: do you need a published breakdown by age group, or are you trying to identify people? Do not use tables to re-identify individuals, or assume that combining small tables is always safe. Useful analysis chooses a level suited to the question and respects access rules.

For reuse of published data, the website’s legal notice permits use of its data with acknowledgement of the source. Check the relevant institution’s terms when using another website or product; do not apply one page’s rule to every Turkish portal. In the discussion itself, identify the table, its period, link and version date so that readers can repeat your steps and examine the figure’s original meaning.

How to assess a chart being shared online

Begin with its title, then examine the axis labels, unit and period. Truncating a vertical axis may make a small difference look large, while a selected end point may hide important changes before or after it. This does not make every abbreviated chart misleading, but calls for checking the underlying figures and context. Look for regularly spaced periods, gaps or changes in measurement.

Next examine the aggregation. A chart covering several areas may conceal differences in population size or service definitions; ranking areas by counts differs from ranking them by rates. Do not divide by an unsuitable denominator merely to produce a percentage. The denominator must relate to what the indicator measures. If the necessary explanation is unavailable, describe the chart accurately rather than constructing a ranking whose meaning cannot be defended.

Finally, distinguish the image publisher’s commentary from the institution’s data. “These figures prove the decision succeeded” may be the publisher’s claim rather than a finding in the release. Return to the original and examine the methodology, period and other relevant indicators. You can accept a figure’s accuracy while disputing its interpretation, or partly accept an interpretation while explaining the limits of its supporting evidence.

A short record for every figure you quote

What is the indicator called, and what does its definition include?
Who produced it, and who is displaying it?
What are the reference period, place and population covered?
What are the unit, denominator and data-collection method?
Is the value provisional, final or a forecast?
Has the series been revised, and when did you save your version?
Does the claim describe the figure, or offer an explanation requiring more evidence?
Can readers follow the link to examine the table and its footnotes?

Keep this record in a personal note for each subject you follow. Enter the details once, then update the version date when a new release appears. If an essential definition is missing, make that gap explicit in your conclusion rather than filling it with guesswork. “There is insufficient detail for this comparison” can be a useful finding that protects the quality of the discussion, rather than an admission that the research has failed.

Frequently asked questions

Can an official figure alone establish a political cause?
It may establish a change within the published definition and period; attributing that change to a particular cause requires an appropriate research design or additional evidence. Look for simultaneous changes, different definitions and alternative explanations. Make clear when an explanation is your own conclusion, and identify its limitations rather than automatically attributing it to the table’s producer.

Why does a page I open in 2026 contain data for 2025?
Publication and access dates differ from the period measured. Record the reference period stated in the table, followed by its publication or revision date where necessary. A recently updated page does not move all its data into the current year, and a report covering an earlier period may be the latest complete result available for a particular subject.

Does a missing figure mean zero?
Do not equate zero with a blank cell or special symbol before reading the footnotes. A value may be unavailable, not applicable or withheld under publication rules; its meaning depends on the table’s explanation. Replacing a blank with zero yourself can change totals and comparisons, creating a trend the source never reported.

Do I need analysis software to read a statistical release?
Start with the title, definition, period, unit and footnotes. If you make a simple calculation, retain the working and denominator, and distinguish hypothetical examples from published results. Software can help organise information, but a sound question and understanding of definitions come first. Even a mathematically correct calculation may support an invalid conclusion if it combines incompatible units.
Related topics

Reading Official Statistics and Open Data in Türkiye: A Practical Guide for 2026

سفير طرابزون

Active member