SDMX Metadata Concepts

Download PDF

0.a. Goal

Goal 16: Promote peaceful and inclusive societies for sustainable development, provide access to justice for all and build effective, accountable and inclusive institutions at all levels

0.b. Target

Target 16.5: Substantially reduce corruption and bribery in all their forms

0.c. Indicator

Indicator 16.5.2: Proportion of businesses that had at least one contact with a public official and that paid a bribe to a public official, or were asked for a bribe by those public officials during the previous 12 months

0.d. Series

Bribery incidence (% of firms experiencing at least one bribe payment request)

0.e. Metadata update

2023-03-31

0.g. International organisations(s) responsible for global monitoring

World Bank (WB)

1.a. Organisation

World Bank (WB)

2.a. Definition and concepts

Definition:

The percent of firms experiencing at least one bribe payment request across 6 public transactions dealing with utilities access, permits, licenses, and taxes.

In every Enterprise Survey (www.enterprisesurveys.org), there are standard questions which ask the survey respondent if they were expected to give a gift or informal payment during a transaction with a public official. There are six, separate transactions which make up this indicator, they include an application for an electrical connection, an application for a water connection, an application for a construction-related permit, an application for an import license, an application for an operating license, and during an inspection/meeting with tax officials. In all of these transactions, if the respondent indicates ‘yes’ they had the transaction (e.g. they applied for an import license), then there is a follow-up question which asks if the respondent was expected to provide a gift or an informal payment during this transaction (an application or meeting). The response options include “yes”, “no”, “don’t know”,and “refuse”. Note that refusals are accepted and recorded but for the purposes of indicator construction, refusals are considered as a ‘yes’. The indicator 16.5.2 is measuring whether the respondent indicated ‘yes’ to a bribe payment for any of these six transactions.

Enterprise Surveys are firm-level surveys conducted in World Bank client countries. The survey focuses on various aspects of the business environment as well as firm’s outcome measures such as annual sales, productivity, etc. The surveys are conducted via face-to-face interviews with the top manager or business owner. For each country, the survey is conducted approximately every 4-5 years.

Concepts:

The respondents to the Enterprise Survey are firms- either manufacturing or services establishments. These are registered (formal) firms with 5+ employees. The firms are either fully or partially private (100% state-owned firms are ineligible for the Enterprise Survey). More information on the survey methodology can be found on the Methodology page of the website: www.enterprisesurveys.org/methodology

A gift or an informal payment is considered a ‘bribe’.

2.b. Unit of measure

Percent (%) of firms experiencing at least one bribe payment request

3.a. Data sources

The website for Enterprise Surveys (www.enterprisesurveys.org) provides all metadata, including survey questionnaires and implementation reports for all Enterprise Surveys. The implementation reports indicate the sample size, sample frame used, dates/duration of fieldwork, the response rates, etc.

Registration to the Enterprise Survey’s website is free and the website’s data portal allows users to access the raw data and survey documentation for each survey.

3.b. Data collection method

The World Bank conducts the Enterprise Surveys in client countries. The surveys are comparable as the survey methodology is applied in a consistent manner across countries: obtaining suitable sample frames, eligibility criteria for respondent firms, survey sample design, core questionnaire elements across every country, standardized QC checks on the received data, standardized computation of sampling weights, etc.

3.c. Data collection calendar

The Surveys are ongoing. Information on current projects can be found at: http://www.enterprisesurveys.org/Methodology/Current-projects

3.d. Data release calendar

The indicators on the Enterprise Surveys website are updated whenever a new survey has been completed and uploaded to the website. For each country, only the most recently completed survey is used when calculating the indicator.

3.e. Data providers

The indicator is derived from Enterprise Surveys which are conducted by the World Bank. The World Bank usually hires a private contractor (typically a market research company) to conduct the survey fieldwork.

3.f. Data compilers

World Bank

3.g. Institutional mandate

The World Bank conducts Enterprise Surveys across the world, but mostly in developing countries. There is an institutional mandate that this data be collected and released for the public good of information. All of the Enterprise Surveys data is anonymized and published on the Enterprise Surveys website. The data can be downloaded, free of charge, for all registered users of the website’s data portal.

4.a. Rationale

The rationale for this indicator is to ascertain whether firms are solicited for gifts or informal payments (i.e. bribes) when undertaking transactions that involve public officials. Applying for regulatory licenses, obtaining utility connections, and paying taxes are required of formal forms in most countries and hence the rational for this indicator is to measure the incidence of corruption during these routine transactions. The key strength of the Enterprise Survey is that most of the questions in the survey pertain to the actual, day-to-day experiences of the firm; these questions regarding corruption are not opinion-based question but rather are grounded in the firm’s day-to-day reality.

4.b. Comment and limitations

The key strength of the Enterprise Survey is that most of the questions in the survey pertain to the actual, day-to-day experiences of the firm; these questions regarding corruption are not opinion-based question but rather are grounded in the firm’s day-to-day reality.

The limitations include that some countries’ data is almost 10 years old (e.g. Burkina Faso and Brazil). This is due to the fact that these face-to-face survey projects can be expensive in some countries and hence due to budget limitations, the World Bank hasn’t been able to update some of the Enterprise Surveys data in a subset of countries. Another limitation is that the surveys are done mostly in World Bank client countries and hence several high-income countries are not covered by the surveys (US, Canada, UK, Singapore, Japan, GCC countries, etc.).

Another limitation may be the sensitive nature of corruption. In some countries/cultures, firms may not be comfortable answering questions on corruption. Although the data is collected under the context of confidentiality, firms may refuse to answer the question if they have been subject to bribery solicitations. Hence, in some countries, the actual incidence of this particular type of corruption may be higher than the calculated indicator value.

4.c. Method of computation

The indicator is calculated for each country, by looking at the proportion of firms which answered ‘yes’ to the survey questions. For all Enterprise Survey projects conducted since 2006, the resulting dataset has sampling weights. Hence the indicator value, which is computed using Stata, incorporates these sampling weights as well as the design strata.

4.d. Validation

This indicator is computed using data collected from the World Bank’s Enterprise Surveys. A detailed manual and guide on the Enterprise Surveys implementation is found here (https://www.enterprisesurveys.org/content/dam/enterprisesurveys/documents/methodology/Enterprise%20Surveys_Manual%20and%20Guide.pdf). Section 4.4 “Data Collection Cycle” of this document describes the processes in place used to validate or check the survey data which is collected to ensure quality.

4.e. Adjustments

For any given survey, during the quality checks outlined in the Enterprise Surveys manual and guide (section 4.4), if inconsistencies or mistakes are found in the data, the World Bank transmits this feedback to the fieldwork team that is conducting the survey in the first place. The fieldwork team should make sure that any data mistakes are corrected (or if the data is indeed correct, provide the justification to the World Bank) when submitting the final survey dataset.

4.f. Treatment of missing values (i) at country level and (ii) at regional level

• At country level

The indicator value is not imputed for countries which do not have an Enterprise Survey.

• At regional and global levels

Regional and global aggregates of the indicator are derived from completed surveys. A single point estimate is created for each country and a global/regional aggregate takes a simple average of every country’s point estimate (when there is available data for that country). For example the East Asia Pacific average (point estimate) for the indicator does not include Japan since there is no Enterprise Survey for Japan.

4.g. Regional aggregations

Regional and global aggregates are computed by taking the simple average of the indicator value for all relevant countries. When producing regional and global aggregates as presented on the Enterprise Surveys website, note that only surveys posted during years 2010 onwards are used.

4.h. Methods and guidance available to countries for the compilation of the data at the national level

We recommend users consult the Enterprise Surveys website to learn about the overall survey methodology and learn which countries are available for benchmarking purposes. http://www.enterprisesurveys.org/methodology

4.i. Quality management

A detailed manual and guide on the Enterprise Surveys implementation is found here (https://www.enterprisesurveys.org/content/dam/enterprisesurveys/documents/methodology/Enterprise%20Surveys_Manual%20and%20Guide.pdf). This manual provides a comprehensive overview of the quality management of the Enterprise Surveys.

4.j. Quality assurance

The process of quality assurance includes the review of survey questionnaires/documentations/metadata, examination of reliability of data, and making sure they comply with international standards (e.g. workforce concepts in the survey questions correspond to International Labour Organization (ILO) standards), and examining the consistency and coherence within the data set as well as with the time series of data and the resulting indicators.

4.k. Quality assessment

When conducting our survey projects, the implementing fieldwork team must send periodic batches of completed interviews to the World Bank so that we can run our own quality control programs on the data. After running these programs, we provide the QC feedback to the implementing fieldwork team so that survey data, which has been flagged, can be verified and continuously improved. This is how we continuously monitor the survey data while the projects are in the field.

The World Bank collects this survey data for the public good of information. For an individual survey project, once the data is collected and considered finalized (after our own internal QC processes), the survey data is published on the World Bank’s Enterprise Surveys website.

5. Data availability and disaggregation

Data availability:

Data for around 154 economies were collected.

Time series:

Surveys are implemented in around 50 countries every year. Data frequency for each country is around 4-5 years.

Disaggregation:

The Enterprise Survey captures several descriptive characteristics of the respondent firms including: gender of top manager, primary business activity of the firm, subnational location of the firm, exporting status, number of employees, degree of foreign ownership, and several other characteristics. Hence the indicator can be disaggregated by the levels of these individual characteristics.

6. Comparability/deviation from international standards

Sources of discrepancies:

We are unaware of any country-produced data on this indicator.