Understanding Skill Gaps in Firms

Annex B. Data availability and use

All countries participating in the employer module have taken steps to ensure the accessibility of their data for scientific research purposes. In addition, the OECD Secretariat has produced a public use file (PUF) to facilitate wider scientific use.

National level data access and use

Ensuring access to survey data is crucial for the advancement of scientific research. Countries have taken different approaches to making the national data of the PIAAC Employer module available, typically together with the CVTS data:

• Hungary: Main results of the CVTS survey are available in the form of static and dynamic tables on the webpage of the HCSO. HCSO facilitates scientific research by opening data files upon request, following an accreditation study. Researchers have controlled access to de‑identified datasets, ensuring legal and methodological guarantees. Access rights for staff are regulated and details of confidentiality are published on the HCSO website.

• Italy provides access to CVTS survey results through the Istat website. Researchers seeking more detailed results can request additional data processing, considering compatibility with the sampling plan. Access to CVTS micro-data is facilitated by the provision of anonymised and perturbed datasets, with priority given to respondents’ privacy. Istat provides micro-data files for Statistical Offices or entities belonging to the National Statistical System.

• Portugal: CVTS results for Portugal have been published on the website of the Office of Strategy and Planning of the Ministry of Labour, Solidarity and Social Security (GEP/MTSSS). Accredited researchers can request access to anonymised micro-data of both the CVTS and the PIAAC Employer Module from the PT NSI, ensuring responsible and controlled access.

• Slovak Republic: Data from the Slovak Republic are accessible through the online database DATAcube, managed by the Statistical Office of the Slovak Republic (SO SR). The platform, which has been operational since 31 October 2022, provides free access to all users without the need for registration. In addition, researchers can request anonymised data for scientific purposes under strict conditions.

• The Netherlands: In the Netherlands, micro-data from the CVTS 2020 survey are accessible to registered users through a remote access environment. Further information on access is available on the Statistics Netherlands website. Although there is no specific release calendar, Statistics Netherlands maintains an overview of available datasets for the authorised public.

International Public Use File

The OECD Secretariat has created a PUF that is freely and publicly available on the OECD website. This dataset brings together and harmonises data from five participating countries (Italy, Hungary, Portugal, the Netherlands and the Slovak Republic) to allow external researchers to access and use the data for their own resource purposes. The final form of the PUF was decided together with the participating countries, balancing concerns about confidentiality and data protection risks with usability for research purposes.

Structure and content of the aggregated dataset

The PUF is presented as an aggregated dataset, removing the micro-data nature of the original enterprise level data. Each row in the dataset represents a combination of key groupings by country, industry, and enterprise size. Specifically, the dataset includes five countries – Italy, Hungary, Portugal, the Netherlands and the Slovak Republic – and is structured by:

• five economic activity classes based on the NACE 5 classification.

• three enterprise size classes: 10‑49 persons employed, 50‑249 persons employed and 250+ persons employed.

For each combination of country, economic activity and enterprise size, the numerical value in the cell indicates the grossed-up total number of enterprises in that category that provided a specific response to the survey. As the dataset reflects the total number of enterprises within each grouping, no further weighting needs to be applied to ensure an accurate representation of the population of enterprises across sectors and size classes within each country.

To protect confidentiality, standard data protection protocols are followed and any (non-weighted) grouping by country and size class with less than 10 enterprises is excluded from the dataset.

The dataset includes both core and optional questions. A codebook is provided below.

Transformations applied

The following transformations have been applied to the micro-dataset:

• Categorical variables: The values of categorical variables are converted into separate variables representing the number of enterprises in each cell reporting a particular value. Where too many cells would need to be suppressed due to low numbers per cell, categories are merged to create larger categories. Population weights are applied to produce weighted counts for all categorical variables.

• New variables: Additional variables are created to improve the usefulness of the dataset, including the number of enterprises per row (i.e. per country, enterprise size and industry group).

• Variable selection: Variables with limited analytical value or a large amount of missing data when aggregated, such as region or main reference market, are excluded from the dataset.

Use and analytical value

The aggregated dataset will allow researchers to perform descriptive analyses based on country, industry, and firm size groupings. While this limits the scope of analysis compared to individual-level data, some advanced analyses can still be performed. Researchers will also be able to link the aggregated data to other external datasets, such as the PIAAC Household Survey or the Continuing Vocational Training Survey (CVTS), provided that these datasets allow the creation of similar sub-groups. Importantly, this option eliminates any risk of identifying individual enterprises, thus ensuring full compliance with countries’ concern around data protection and confidentiality.

International codebook

Table A B.1. Employer Module: International codebook

Variable Name	Values	Description	Related Survey Question
country	HU IT NL PT SK	Country code	n/a
firm_size	10‑49 employees 50‑249 employees 250+ employees	Firm size category	Q8
industry	NACE Rev. 2: B, C, D, E NACE Rev. 2: F NACE Rev. 2: G, H, I NACE Rev. 2: L, M, N, R, S NACE Rev. 2; J, K	Industry NACE 5 category	Q7
refyear	2020	Reference year	n/a
n	Numeric	Count of firms in row	n/a
birthyear_1	Numeric	Count of firms created before 2000 (inclusive)	Q10
birthyear_2	Numeric	Count of firms created between 2001 and 2020 (inclusive)	Q10
d_emp_1	Numeric	Count of firms that experienced an increase in headcount in reference year	Q9
d_emp_2	Numeric	Count of firms that experienced a decrease in headcount in reference year	Q9
d_emp_3	Numeric	Count of firms where headcount in reference year stayed more or less the same	Q9
gap_0	Numeric	Count of firms with no skill gap or don’t know	Q1 (None or Don’t know)
gap_1	Numeric	Count of firms with a skill gap	Q1 (Few, Some, Most or All)
under_flag_1	Numeric	Count of firms that have identified at least one area of skill gap	Q2
under_a_1	Numeric	Count of firms that have identified skill gap in: General IT skills	Q2
under_b_1	Numeric	Count of firms that have identified skill gap in: IT professional skills	Q2
under_c_1	Numeric	Count of firms that have identified skill gap in: Management skills	Q2
under_d_1	Numeric	Count of firms that have identified skill gap in: Team working skills	Q2
under_e_1	Numeric	Count of firms that have identified skill gap in: Customer handling skills	Q2
under_f_1	Numeric	Count of firms that have identified skill gap in: Office administration skills	Q2
under_g_1	Numeric	Count of firms that have identified skill gap in: Foreign language skills	Q2
under_h_1	Numeric	Count of firms that have identified skill gap in: Technical, practical or job specific skills	Q2
under_i_1	Numeric	Count of firms that have identified skill gap in: Oral or written communication skills	Q2
under_j_1	Numeric	Count of firms that have identified skill gap in: Mathematics or calculating skills	Q2
under_k_1	Numeric	Count of firms that have identified skill gap in: Reading skills	Q2
under_l_1	Numeric	Count of firms that have identified skill gap in: Problem solving skills	Q2
under_m_1	Numeric	Count of firms that have identified skill gap in: Other	Q2
solution_flag_1	Numeric	Count of firms that have identified at least one action taken to alleviate skill gap	Q3
solution_a_1	Numeric	Count of firms that enacted following solution: Provide training	Q3
solution_b_1	Numeric	Count of firms that enacted following solution: Offer internal job mobility	Q3
solution_c_1	Numeric	Count of firms that enacted following solution: Recruit new staff with suitable qualifications, skills and competencies	Q3
solution_d_1	Numeric	Count of firms that enacted following solution: Recruit new staff combined with specific training	Q3
solution_e_1	Numeric	Count of firms that enacted following solution: Implement mentoring /or buddying scheme	Q3
solution_f_1	Numeric	Solution: Increase performance monitoring	Q3
solution_g_1	Numeric	Count of firms that enacted following solution: Provide feedback to staff	Q3
solution_h_1	Numeric	Count of firms that enacted following solution: Change work practices	Q3
solution_i_1	Numeric	Count of firms that enacted following solution: Reallocate work	Q3
solution_j_1	Numeric	Count of firms that enacted following solution: Automate production	Q3
solution_k_1	Numeric	Count of firms that enacted following solution: Implement domestic or foreign outsourcing	Q3
solution_l_1	Numeric	Count of firms that enacted following solution: Abandon the activity	Q3
change_flag_1	Numeric	Count of firms where at least one change has significantly affected enterprise in previous 3 years	Q4
change_a_1	Numeric	Count of firms where there has been a change to: Machinery	Q4
change_b_1	Numeric	Count of firms where there has been a change to: Information and communication technologies and processes	Q4
change_c_1	Numeric	Count of firms where there has been a change to: Working methods and organisational practices	Q4
change_d_1	Numeric	Count of firms where there has been a change to: Domestic outsourcing practices	Q4
change_e_1	Numeric	Count of firms where there has been a change to: Foreign outsourcing practices	Q4
change_f_1	Numeric	Count of firms where there has been a change to: Products or services	Q4
change_g_1	Numeric	Count of firms where there has been a change to: Amount of contact with clients or customers	Q4
trainchange_1	Numeric	Count of firms where training was offered to address above changes	Q5 (any Yes response)
inno_1	Numeric	Count of firms that innovate very rarely or rarely	QA2 (Very rarely or rarely)
inno_2	Numeric	Count of firms that innovate occasionally	QA2 (Occasionally)
inno_3	Numeric	Count of firms that innovate very frequently or frequently	QA2 (Very frequently or frequently)
quality_1	Numeric	Count of firms that compete in a market with basic quality goods/services	QA3 (1, 2 or 3 on scale)
quality_2	Numeric	Count of firms that compete in a market with premium quality goods/services	QA3 (4 or 5 on scale)
saa_0	Numeric	Count of firms that do not conduct skills assessment and anticipation	QB1
saa_1	Numeric	Count of firms that conduct skills assessment and anticipation on an ad hoc basis	QB1
saa_2	Numeric	Count of firms that conduct skills assessment and anticipation regularly	QB1
effect_flag_1	Numeric	Count of firms that have experienced at least one effect of having a skill gap	QA1
effect_a_1	Numeric	Count of firms that have experienced following effect: Not able to take on as much business as you would like	QA1
effect_b_1	Numeric	Count of firms that have experienced following effect: Loss of business or orders to competitors	QA1
effect_c_1	Numeric	Count of firms that have experienced following effect: Delays in developing new products or services	QA1
effect_d_1	Numeric	Effect: Difficulty in meeting quality standards	QA1
effect_e_1	Numeric	Count of firms that have experienced following effect: Increased operating costs	QA1
effect_f_1	Numeric	Count of firms that have experienced following effect: Difficulty in introducing new working practices	QA1
effect_g_1	Numeric	Count of firms that have experienced following effect: Increased workload for existing staff	QA1
effect_h_1	Numeric	Count of firms that have experienced following effect: Difficulties in meeting customer service objectives	QA1
effect_i_1	Numeric	Count of firms that have experienced following effect: The withdrawal of certain products or services altogether	QA1
effect_j_1	Numeric	Count of firms that have experienced following effect: Difficulties in introducing technological change	QA1
train_flag_1	Numeric	Count of firms that have provided at least one training activity in reference year	QB2
train_a_1	Numeric	Count of firms that have provided: Internal CVT courses	QB2
train_b_1	Numeric	Count of firms that have provided: External CVT courses	QB2
train_c_1	Numeric	Count of firms that have provided: Guided-on-the‑job training	QB2
train_d_1	Numeric	Count of firms that have provided: Job rotation, exchanges, secondments or study visits	QB2
train_e_1	Numeric	Count of firms that have provided: Conferences, workshops, trade fairs or lectures	QB2
train_f_1	Numeric	Count of firms that have provided: Learning or quality circles	QB2
train_g_1	Numeric	Count of firms that have provided: Self-directed learning/e‑learning	QB2
lowtrain_a_1	Numeric	Count of firms that have identified following reason for limited training: Existing qualifications, skills and competences of current staff were appropriate to current needs of enterprise	QB4
lowtrain_b_1	Numeric	Count of firms that have identified following reason for limited training: Preferred strategy was to recruit individuals with the required qualifications, skills and competences	QB4
lowtrain_c_1	Numeric	Why limited training: Difficulties assessing training needs in the enterprise	QB4
lowtrain_d_1	Numeric	Count of firms that have identified following reason for limited training: Lack of suitable offers of CVT courses in the market	QB4
lowtrain_e_1	Numeric	Count of firms that have identified following reason for limited training: High costs of CVT courses	QB4
lowtrain_f_1	Numeric	Count of firms that have identified following reason for limited training: Higher focus on IVT provision than on CVT	QB4
lowtrain_g_1	Numeric	Count of firms that have identified following reason for limited training: Major efforts in CVT made in recent years	QB4
lowtrain_h_1	Numeric	Count of firms that have identified following reason for limited training: High workload and no time available for staff to participate in CVT	QB4
lowtrain_i_1	Numeric	Count of firms that have identified following reason for limited training: Other reasons	QB4
notrain_a_1	Numeric	Count of firms that have identified following reason for no training: Existing qualifications, skills and competences of current staff were appropriate to current needs of enterprise	QB4
notrain_b_1	Numeric	Count of firms that have identified following reason for no training: Preferred strategy was to recruit individuals with the required qualifications, skills and competences	QB4
notrain_c_1	Numeric	Count of firms that have identified following reason for no training: Difficulties assessing training needs in the enterprise	QB4
notrain_d_1	Numeric	Count of firms that have identified following reason for no training: Lack of suitable offers of CVT courses in the market	QB4
notrain_e_1	Numeric	Count of firms that have identified following reason for no training: High costs of CVT courses	QB4
notrain_f_1	Numeric	Count of firms that have identified following reason for no training: Higher focus on IVT provision than on CVT	QB4
notrain_g_1	Numeric	Count of firms that have identified following reason for no training: Major efforts in CVT made in recent years	QB4
notrain_h_1	Numeric	Count of firms that have identified following reason for no training: High workload and no time available for staff to participate in CVT	QB4
notrain_i_1	Numeric	Count of firms that have identified following reason for no training: Other reasons	QB4
hardhire_a_1	Numeric	Count of firms that experienced difficulty recruiting staff for jobs which normally require a formal vocational qualification	QD1 (any Yes response)
hardhire_b_1	Numeric	Count of firms that experienced difficulty recruiting staff for jobs which normally require a university degree	QD1 (any Yes response)
hardhire_c_1	Numeric	Count of firms that experienced difficulty recruiting staff for jobs that do not require any formal qualification nor degree	QD1 (any Yes response)
hardkeep_1	Numeric	Count of firms that experienced difficulty retaining employed staff	QD2
hardrecruit_1	Numeric	Count of firms that experienced difficulty recruiting skilled candidates	QD3
underhire_flag_1	Numeric	Count of firms that have identified at least one skill area for which it was difficult to hire skilled candidates	QD4
underhire_a_1	Numeric	Count of firms with difficulty hiring skilled candidates in: General IT skills	QD4
underhire_b_1	Numeric	Count of firms with difficulty hiring skilled candidates in: IT professional skills	QD4
underhire_c_1	Numeric	Count of firms with difficulty hiring skilled candidates in: Management skills	QD4
underhire_d_1	Numeric	Count of firms with difficulty hiring skilled candidates in: Team working skills	QD4
underhire_e_1	Numeric	Count of firms with difficulty hiring skilled candidates in: Customer handling skills	QD4
underhire_f_1	Numeric	Count of firms with difficulty hiring skilled candidates in: Office administration skills	QD4
underhire_g_1	Numeric	Count of firms with difficulty hiring skilled candidates in: Foreign language skills	QD4
underhire_h_1	Numeric	Count of firms with difficulty hiring skilled candidates in: Technical, practical or job specific skills	QD4
underhire_i_1	Numeric	Count of firms with difficulty hiring skilled candidates in: Oral or written communication skills	QD4
underhire_j_1	Numeric	Count of firms with difficulty hiring skilled candidates in: Mathematics or calculating skills	QD4
underhire_k_1	Numeric	Count of firms with difficulty hiring skilled candidates in: Reading skills	QD4
underhire_l_1	Numeric	Count of firms with difficulty hiring skilled candidates in: Problem solving skills	QD4
underhire_m_1	Numeric	Count of firms with difficulty hiring skilled candidates in: Other	QD4
causehire_a_1	Numeric	Count of firms that experienced difficulty hiring for following reason: Wages are lower than in other organisations	QD5
causehire_b_1	Numeric	Count of firms that experienced difficulty hiring for following reason: Geographic location	QD5
causehire_c_1	Numeric	Count of firms that experienced difficulty hiring for following reason: Unattractive conditions of employment	QD5
causehire_d_1	Numeric	Count of firms that experienced difficulty hiring for following reason: Lack of career progression	QD5
causehire_e_1	Numeric	Count of firms that experienced difficulty hiring for following reason: Long/unsocial hours	QD5
causehire_f_1	Numeric	Count of firms that experienced difficulty hiring for following reason: High competition from other employers	QD5
causehire_g_1	Numeric	Count of firms that experienced difficulty hiring for following reason: Not enough people interested in doing type of work	QD5
causehire_h_1	Numeric	Count of firms that experienced difficulty hiring for following reason: Staff don’t want long term commitment	QD5
causehire_i_1	Numeric	Count of firms that experienced difficulty hiring for following reason: Other reasons	QD5
hrhead	Numeric	Count of firms where person in charge of human resources reports directly to head of enterprise	QC1
pcteam	Numeric	Average percentage of employed staff working in teams in row	QC2
pcmeet	Numeric	Average percentage of employed staff meeting to improve practices in row	QC2
pcdatabase	Numeric	Average percentage of employed staff updating database of good practices in row	QC2

Note: Population weights have been applied to all numeric variables.

Source: OECD international codebook created for public use purposes. Based off data submitted by countries.