The time that, according to EPA, has dropped substantially after the pandemic. This study argues that these results are largely influenced by the methodological change introduced in 2021 , which affects the measurement of time not worked. In the new EPA wondered directly by the time not worked due to the holiday or disease and, by extension, generates a fall of the effective working hours. The comparison of the new EPA with the ETCL reveals that levels of hours worked by not vacation and sick are now more plausible. This conclusion is reinforced by observing that, until 2020 effective, annual full-time employees always exceeded the legal maximum, a scenario hard to believe.

Keywords:working time, working hours, holidays, sick leave

According to the LFS, effective working time in Spain has significantly decreased following the pandemic. This study argues that this change is great influenced by the methodological change implemented in the LFS in 2021 , which affects the measurement of unworked time. The new LFS records more accurately unworked time due to vacations or illness, leading to a decrease in the effective working hours. Comparing the new series with the ETCL reveals that the levels of unworked time due to vacations and illness now appear more plausible. This conclusion is reinforced by the observation that, until 2020 , the effective annual working hours of full-time employees always exceeded the legal maximum, a scenario that is difficult to believe.

Keywords:working time, workday, vacations, sick leave

The extent of working time is a key issue for two reasons. Firstly, because it provides for the inputcritical to measure the contribution of the labour factor of economic output. The use, which generally has been measured in terms of people in work or jobs, is complemented by a more precise measurement time actually worked, expressed in the total number of hours worked. As a result, the accuracy in the calculation of working hours has a direct impact on labour productivity. Secondly, working time is a key element in the design and effectiveness of various social policies that regulate working hours. These policies seeking to achieve a maximum of hours (as stipulated in the status of workers and collective agreements) and to secure other labour rights, as a vacation, sick leave and leave for childcare, among others. Therefore, the correct measurement of the working day gives us the best indicator on the effective enjoyment of these rights.

The active population survey (EPA) is one of the main sources for analysing the situation of employment and unemployment in Spain. This survey also collects information on key aspects such as the demographic and social workers, and also about working conditions. As part of this Last block is measured working time, which are classified in terms of both time and time-effective in one particular week. The difference between the meter is crucial to reflect the fluctuations and circumstances affecting the time actually worked for the territories. These differences may be due to various factors, both those that can be considered as a predictable (for example, holidays or seasonal changes in the day) with respect to the unexpected or not programmed (ERTEs, sick leave or accident).

In Spain, since the beginning of the pandemic in 2020 the EPA collects a major difference between normal day and the effective (see graph 1 ). The fall in the actual average has received significant attention from the main institutions dedicated to the economic analysis (Botelho, Consolo and Da Silva 2021 Square; 2023 ; Astinova et al. 2024 ). Although these studies do not have come to establish consensus on its causes, it is possible to drop some basic scenarios. The growing gap between actual hours and habitual indicates that the fall in the effective day cannot be attributed to a reduction in the standard working day agreed, which is usually very close to the usual hours. It cannot be attributed to one important change in the relative weight of the full-time employment and employment. In fact, the main trends pointing in opposite direction since the rate of full-time workers has increased slightly. Accordingly, the leading candidate to explain the reduction in the time that lies in the increase of time does not worked.

In this work argumentaré need to improve the indicators to make a full analysis of variances between standard time and effectively. First, to measure the working time, it should be given equal attention to the full and partial absences, i.e., both of those employees who had not worked any day of the week as those employees who, even to work, make it less than usual in their journey. Secondly, to compare both groups (i.e., those in the absence in whole or in part), should standardize the measurement of time not worked in terms of days not worked or hours. Thirdly, that when taking into account these two indicators showed thatthe methodological change introduced in the EPA in 2021 causing a breakdown in the results to more accurately measured absences for vacation and sick.

This exercise indicates that the bulk of the difference is due to a change of a questionnaire on the other, and which the following results: First, the new EPA wondered all interviewed it had been absent due to holidays, due to illness and for other reasons. Enter these three successive questions increases dramatically the percentage of employees who report to be absent for these reasons, especially of those who are on holiday. this effect, coupled with slight variations in the average time that had been absent employees who had not worked, enables a 40 % of the decline observed in the actual average of work relates to methodological reasons. The remainder of the increase of time does not worked due to the increase in the ratio of workers in complete absence for the same two reasons (holidays or illness), a change that principle cannot be attributed to a change in the question, but that certainly requires that all users intensified precautions when using the series of the EPA.

The second most important result is obtained by comparing time worked in the EPA for vacation and sick with another baseline survey Quarterly, the survey of Labour Cost (ETCL), addressed to companies. This comparison shows that up 2020 the EPA infraestimaba periods not worked for holidays and sick, and from new guest 2021 levels and trends in both registries (EPA and ETCL) coincide to a greater extent. These results suggest that the effective day under the EPA 2020 so far has been overstated 2020 , a point that is confirmed in the accounting of actual time of full-time employees. The actual average time of this group, in annual count, stood consistently above the statutory maximum established in the status of workers ( 1 . 826 p.m.). This incongruity, again, should compel all users are having the precautions to interpret the time of the working group until the methodological change 2021 and, moreover, to derive any conclusion on the change in working time after the pandemic. In conclusion, the EPA has improved its procedures for measuring the effective working hours and previous records showed an upward bias. This change is also open a new perspective if we want to make a comparison.

The article is organized as follows: To explain the meaning of this change, in the following section discusses how was measured working time in the EPA prior to the change of methodology and the innovations introduced in 2021 I. After the key results in terms of percentage of workers who report to be absent from their work, either for the whole week (complete absence), or only for a more limited period (partial absence). next, i examine the consistency of results, considering the time worked in terms of hours or days not worked for the holidays, and compared to other sources (mainly the ETCL) and the day after the full-time employees. The working group concludes with a number of possible lines of research and recommendations to establish good practices.

The EPA is one of the most complex surveys that exist in our country. Each interview about 60 . 000 homes and collected first-hand information on 140 . 000 people, so that results provide a significant level of detail. Their large size is added the complexity of the questionnaire of the EPA, in the case of workers, incorporates until almost 50 questions on a broad range of issues on its use. Among this set of questions, one of the priority areas is the measure of working time and absences for various reasons.

The investigation requires that the survey questions will give rise to clear answers and unbiased interviewees. To analyse the accuracy of these questions is necessary to decide how the EPA during the period of 2005 to 2020 , where there was a model questionnaire (national institute of statistics 2005 ), with respect for the period that begins in 2021 when it introduces a new one (national institute of statistics 2021 ). Figure 2 reflects the main features of these two models of the questionnaire in connection with the working day. As noted at first sight, both are striking similarities in questions relating to the usual time and the effective, but also in the recording of complete absences and overtime. However, differ in how to measure the partial absences due to holidays, illness or other causes. this effect is even greater if it serves the sequential logic of the questionnaire.

In both questionnaires of the EPA there is a first block B that seeks to identify the workers and assess the existence of a job, even if they had worked on it during that week. In the guest 2005 the key question was as follows:

B. 4 . The reference week Was a job or business, but not work in This week?

Below, it is asking for the reason to explain this absence, but more important, possible options will not be read in first instance:

B. 5 . What is the main reason why he did not work in this employment? (Not read the options should be noted that the interviewee. If you say several, it requests the main. Only if said that he did not know, read the options. If you still do not know, this option).

The next block (C) of the questionnaire is dedicated to the characteristics of employment. In one part of the interview was announced to you to ask about “ its usual working hours, on its day in a given week and if this week made overtime ”. In the guest 2005 the first question related to the usual time:

V. 23 . How Many hours per week works regularly in this employment? Did not take into account the time to eat.

Followed by a question about whether there had been an increase of the day because of the development of overtime (V. 24 ), and then ask about the actual day in a week:

V. 25 . In the reference week, How many hours worked in this employment? Did not take into account the time to eat

Finally, in the guest 2005 , only to the territories which made positive actual hours (greater than zero) and other than the usual, it was an additional question:

V. 29 . What is the main reason why worked a different number of hours of the usual? (Not read the options. The interviewer checked, as appropriate, according to the response of the informant).

From 2021 it is setting up a new questionnaire on the EPA. This change was born two years earlier with the European regulation 2019 / 1700 and of the implementing regulation ( 2019 / 2240 ) of the commission, which details the variables and questions of the EPA at european level, leading to the inclusion of three new questions:

1. Days of absence of employment for holidays and permits [ABSHOLID]
2. Days of absence of the main employment due to illness, injury or temporary disability own [ABSILLINJ]
3. Days of absence of the main employment for other reasons [ABSOTHER]

The definition and direction of these new questions reflected in the Eurostat methodology document (Eurostat 2021 ).

In spain, the new questionnaire of the EPA implemented in 2021 retrieves methodological changes. The new naq peddled much common questions to the old model, which introduced some new and eliminated others. At its first block, the new EPA remained the same pattern and drawn up by asking the territories that had been absent from their jobs and the ground. Also, in the block devoted to the characteristics of employment started asking about normal working week (equal to the previous V. 23 ), but then adds a new series of questions relating to absences due to holidays:

V. 37 . In the reference week, was absent for holidays, weekends, summer time, flexible?
V. 38 May specify the grounds of absence. I read the options: 1 Holiday. 2 Public holidays. 3 Summer workshop, variable or flexible hours or similar.
V. 39 How many days was absent for such reasons, during the reference week?

At this point, it is noteworthy that the methodological guide clarifies that Eurostat should only be possible to compute accrued vacation compensation as those that are typically had worked.

The guest the EPA introduced questions very similar to pick up time worked for reasons of health or disease. The questions are:

V. 40 . In the reference week, was missing from this employment by health problems, sickness, accident or temporary disability?
V. 41 How many days (or time) was absent for any of these reasons, during the reference week?

Below, another question to detect absences for other reasons, but without mention them. only in completing these questions when it returned to the logic of the previous form, asking first for the extra hours and, at the end, for the effective day:

V. 46 . In the reference week, How many hours worked in this employment? Did not take into account the time to eat

The question that existed in the previous guest the EPA on the difference between the effective and disappears, surely due to be redundant with new queries.

In sum, the new questionnaire of the EPA implemented in 2021 introducing profound changes in the measurement of working time. Since then, in the EPA is explicit in a number of questions the employee has been absent for holidays, illness or other reasons, even though this absence if for a short time (for example, a day). In Addition, the question of effective working hours, which is the keystone of vault of all economic analysis, only takes place after each of these questions, for which necessarily the interviewee is likely to respond with greater precision than in the previous questionnaire.

In the preceding paragraph has been stressed that workers may be absent the whole week (complete absence) or partial if, having worked, what made less than usual. Furthermore, it has been pointed out that the new questionnaire of the EPA question explicitly by absence due to holidays or disease. With a view to gauge the magnitude of change, in this section i examine the magnitude of the absences for both reasons as a result of using the methodological change microdata EPA for employees (National institute of statistics 2025 ). With microdata give the following unpublished data: the percentage of workers in the absence in part by both causes and not the time worked. The Latter variable register for the entire period in terms of hours and, since the change methodological 2021 also, in days not worked.

Figure 3 shows the percentage of wage-earners who answered affirmatively and compared to the records according to the previous question. Please note that the methodological change leads to an increase of approximately 4 points in the percentage of employees in partial absence.

A similar situation applies to workers who, having worked in the reference week were absent due to over a period of time (at least one hour) by reason of sickness or temporary disability. Note (figure 4 ) that change is much lower (an increase of one point), but also that the previous series starting from extremely low levels (around 0 , 2 per cent) and showed a seasonal pattern much less accused with the new methodology.

This methodological change also affects the calculation of time does not worked for both reasons. For this calculation should be borne in mind the options in each period. In the old guest the EPA, time is not working alone could be measured as the difference between the hours and the actual hours. With the new naq this option is still present, but, in addition, the EPA provides information on not worked on holiday and disease.

Figure 5 contains information in terms of recent days not worked. The records confirm the intuition that complete absences are tantamount to a period of absence during 5 days on average. The partial absence, by contrast, are on average around 1 , 6 or 1 , 7 days a week. Figure 6 it allows for the period between the pre- and post-methodological change using the hours. The average for the complete absences at around 36 hours per week, which is a level almost identical to the common to all-time employees. In the absence in part, there are two interesting events. First, that the average length of the partial absence through illness drops drastically environment 17 hours per week to 11 p.m. This situation reflects the increase of workers registered with disease increases the weight of the casualties of shorter duration. In contrast, in absence for holidays the difference is not so remarkable.

Computing periods not worked has the advantage of facilitating comparison with other records, mainly the ETCL. In this statistic is measured the difference between the hours paid and the effective according to various cases, including, holidays and the low for temporary disability. The comparison is feasible once made some small adjustments. It should also be taken into account that the EPA revolves on the main employment of persons interviewed, but it does not include detailed information on ancillary employment The ETCL, however, to investigate, in aggregate form, to all wage earners and self-employed partners to the account of contributions, for which there existed an obligation to contribute at least one day during the month of reference, in any form of contract and working hours. This means that includes workers who are in the company as a second job. This fact, together with the slight difference in the sectoral coverage (the ETCL excludes agriculture and domestic service), makes the results can never be identical.

Figure 7 shows the computation of the hours worked, expressed as average wage earners, according to both registries. At first sight, leads to two conclusions. First, that time has not worked for holidays increases in the EPA in 2021 in approximately 40 %. This magnitude is difficult to explain a change in the current legal framework, either in the status of workers or collective agreements. Please note that the ETCL does not show an increase in the vacation time, a relevant fact for any further interpretation. The second conclusion is that until 2021 the EPA registered levels far below those of the ETCL, whereas since then the opposite is happening, albeit with minor differences. The situation prior to 2021 it is particularly relevant, since what may seem a small difference in their weekly working hours, in reality hides, a significant impact on the annual count. For example, if the gap in these years is about 0 , 7 hours per week, in the annual count a difference of one week typical of labour between the two registers ( 0 , 7 * 52 = 36 , 4 p.m.).

At the time worked by disease a similar situation occurs (figure 8 ) So Far. 2020 the EPA, leaving always levels below the ETCL and partially meet that year, in particular during the period of the former state of alarm (second quarter of 2020 ), Since Then. 2021 in future, both series show a remarkable similarity.

The above results invite you to think that until 2020 the EPA infraestimaba time worked and sobreestimaba effective hours weekly. The following subparagraph seeks to bring the impact of the methodological change in the main series.

If he returns to the sequence of questions of the EPA, can be summed up the computation of the actual hours (I) as a result of the difference between normal hours (HH), hours worked per absences (HAS) and, in addition, the amount of overtime (HX):

I = HH - + HX( 1 )

For the purpose of this work, it is necessary to differentiate between absence due to holidays or period of illness, and in complete absences (HAC) and one-sided ones (PAHS). All other absences (ERTE, relatives or minors, etc.) are included in an additional concept (HAO):

I = HH - HAC - PAHS - HAO + HX( 2 )

With a view to simplifying the presentation is grouped into a final error (Ε) the combined effect of absence for other reasons (ERTE, relatives or minors, etc.), overtime and other factors that in the EPA may give rise to small inconsistencies (non-response, people in a week and individuals on sick leave, etc.):

I = HH - HAC - PAHS + Ε( 3 )

Finally, if you want to explore this mechanism in per capita terms, it suffices to divide each term by the working population (N) and take into account the burden of the collective partial absence (P ₁ andP ₂ ).

I/N = HH/N - P 1 /N HAC/ P 1 - P 2 /N PAHS/P 2 + Ε( 4 )

The breakdown is very simple and seeks only to weigh the effect of the change in the percentage of workers who are in the absence in whole or in part by the two cases. Not seeks to explain all the factors that determine the difference between regular and effective.

The table 1 summarized the major steps in this Last period. It Should Be repaired on some facts. First, as highlighted in the introduction, the usual hours are not to fall in this period (Torrejón et al. 2024 ), forcing a discard other factors not related to periods of absence from work. Secondly, that time has not worked due to vacation and sick explains most of this difference, with the exception of the year 2020 and 2021 when ERTEs it played an outstanding. Thirdly, to break down the impact of absence for these two reasons, it is noted that partial absence have a considerable impact, adding 0 , 7 hours per week at the time worked (or a 38 per cent of the increase in time worked).

The details of these calculations, which is presented in the table 2 , developing a fourth conclusion. In the EPA, the calculation of time does not worked depends crucially on the percentage of workers who claim to be absent from a complete or partial satisfaction of their jobs. The climb so strong of workers in the latter situation, from 2021 – increases 5 points – gives account of the importance of a methodological change that, from another perspective, it may seem trivial. These results are the same way as other recent work (King of the castle 2024 , 12 ), which examines how climate methodological 2021 led to a higher 3 , 6 points in time worked according to the EPA.

Another perspective to address the methodological change is to take the figures for the effective day and compared with the current legal framework. As noted (figure 1 table 2 ), these ranged around 32 , 5 hours per week until 2019 and then go down to an average level of 31 hours 2020 in the future. However, these figures relate to all wage earners, including those full- and part-time. A more precise measure for comparison with the legal maximum working week is to take only to the full-time employees and differentiated between the main branches of activity. The average hours a week to be returned as the average of the four quarters of the year you will then anualizar (multiplying by 52 ). This measure is consistent because the EPA is distributed equally interviews during the weeks of the year, the annual average correctly reflects changes seasons in the working time.

Figure 9 shows these results and includes the statutory maximum established in the current labour legislation ( 1 . 826 hours 27 minutes per year) ¹ . The trends are extremely revealing, since until 2019 all but one sector (public administration, health and education) were by far above the maximum length legal in annual count. Even this exception in the public services sector, health and education is not due to the weight of public employees, to whom the legal maximum working week is not 1 . 826 p.m.; 1 . 642 (equivalent 37 , 5 hours per week and an increased number of annual or permission). Therefore, before 2020 in all sectors were effective average days in annual count, were above the determined by law. Later, if done abstraction of the fall in the average effective day during a pandemic, 2022 in future all of these same sectors is clearly below this threshold, except for one ranging around this figure (agriculture, livestock and fisheries). From another perspective it is noteworthy that the greatest decline of hours worked by employees full time there has been in trade, transport and hotels (would work on average 130 hours less per year), industry ( 127 hours less), construction ( 124 p.m.) and agriculture ( 120 p.m.). The rest of the branches of the service sector have experienced, in theory, decreases less delivered (around 100 hours per year, on average).

However, all these results are affected by this methodological change, so their value is compromised. Climate scenarios to another explained Solely by the improved registration of the absences, both as partial, complete and suggests that, until 2020 , time has not worked in the EPA was not consistent with the evidence collected by other surveys, such as the ETCL, neither the current legal framework.

This article documented that much of the reduction in time of actual work has been in the EPA in recent years has been the result of methodological change introduced in 2021 . Since that time, the guest the EPA question explicitly by absence due to vacations and periods of disease, with the measurement of time not worked is now much more precise. The actual working days previously sobreestimadas and now close to the measures for other records obtained (mainly the ETCL). There is a methodological change should be noted that as such by INE and taken into account by users of the EPA.

These results are also several implications. The first relates to the possibility of establishing a more rigorous comparative exercise. To date, the specialized literature has always sought to compare the actual hours worked between countries, but this task has always faced the challenge to make comparable records and definitions that are different, more so when an annual count. That is why most comparisons tend to use normal hours because these are not affected by the periods not worked due to holidays, holidays, sick leave and other causes. But even today, a notice in Eurostat comprehensive note (Eurostat 2024 a) these problems:

Following the implementation of the regulation in force from 1 January 2021 and the related questionnaire, data harmonization across Member States is still work in progress to ensure data consistency and comparability. Until further notice, Eurostat recommends not using the figures relating to actual working hours for international comparisons.

In addition, the view or download the series of Eurostat data, a footnote to table that recalls the existence of a break in the series in 2021 (“ break in time series ”) in all countries, including spain.

Once you are nodding this methodological change there is reasonable suspicion that there will be a new impetus to be undertaken and calibrate this exercise compared. To date, institutions dedicated to the economic analysis have been placed on the reduction in hours worked per worker territory in all countries and sought differences between exposure to COVID crisis or other factors. However, as indicated in this article, it is highly doubtful that in spain and the rest of the countries of the union has been a fall in the actual hours as strong as the records derived from household surveys. A break in the number, such as the one which occurred in 2021 does not allow for a comparison on a solid foundation for the previous period and subsequent to the pandemic.

Instead, what it could be considered is that the new questionnaire itself better EPA-LFS approaching levels in the actual working time between countries of the union and, especially, absence due to the above-mentioned factors. For example, the graph 10 reflects the situation in 2023 between absences comprehensive and effective day. In particular, the horizontal axis shows the percentage of the territories that, while maintaining its relationship with employment, have not worked in their employment. The vertical axis gives, in contrast, the share of employed population, having worked in the week, he followed 40 or more hours. The comparison confirms the intuition that the countries of northern Europe have effective days shorter (hence the lowest proportion of employed persons working less than 40 p.m.), but also a higher percentage of workers absent from work week (mainly due to holidays, but also for other reasons). The countries of eastern Europe, instead have the opposite situation (lower levels of complete absence, the more effective days). Spain is a compromise, but closer to the levels observed in the northern countries.

Before proceeding to give an explanation of this phenomenon macroeconomic, what is important is to stress the methodological consistency of these figures. Absences due to holidays, holidays and other grounds are a factor differentiating, because even opting for a narrow definition (the complete absence) shows up 12 points between countries according to one extreme to another. The second implication is that the effective implementation of those who work directly affected by these same factors (holidays or sick leave in the absence in part) and hence the strong variations in the percentage of employed workers 40 or more hours per week. Of course, different legal frameworks, sectoral differences or the relative weight of part-time also explain the disparities observed in the usual time or effective. However, the fact is that to date, economists have emphasized more recent factors and not the periods of absence which now it is more accurately. Look At these absences within a generic concept of “ truancy ” would be a further diagnostic error.

This work also also has important implications for the analysis of trends. In Spain, the crisis of the pandemic was followed by a rapid recovery 2021 - 2022 and, later, for the further growth of the spanish economy towards neighbouring countries in the following years. This context, however, has been marked by repeated revisions of the series of national accounting, that showed a much weaker growth. One of the reasons for refusal are used frequently at the time to explain this weak economic growth was depressed levels of hours worked with the increased force which indicated the employment figures (Arellano, Doménech and García 2024 ). Currently, these macroeconomic series have been revised, but many questions remained unanswered and, among them, the paradoxical evolution of productivity (Artola Maynar Melis And white 2024 ). If measured labour productivity in terms of hours worked, the performance of the spanish economy during the period 2019 - 2024 leaving a balance sheet quite optimistic. In contrast, measured by taking the number of employed persons or the number of jobs full time equivalents, the result is much more pessimistic. A review of the working hours would make any serious effort to correct all the estimates and, probably would require further review of national accounts.

Arellano, F. Alfonso, Rafael Doménech, and Juan Ramón García. 2024 “ Wage Growth and recovery of the labour market After The Covid- 19 . ”Mediterranean economic,no. 38 , 113 – 30 .

White Artola, Miguel, and Francisco Melis Maynar. 2024 . “ Productivity after the pandemic: we know less than we think. ”Nothing is free(blog). November 29 , 2024 .https :// nadaesgratis.es/admin/la-productividad-tras-la-pandemia-sabemos-menos-de-lo-que-pensamos-i.

Astinova, Diva, Romain Duval, Niels-Jakob Hansen, Ben Park, mr. Ippei Shibata, and Frederik Toscani. 2024 . “ Dissecting the Decline in Average Hours Worked in Europe. ” IMF Working Paper 24 / 2 .https :// doi.org/ 10 . 5089 / 9798400261183 . 001 .

Botelho, basque, Agostino Consolo, and antonio Dias Da Silva. 2021 . “ Hours Worked in the Euro Area. ” ECB Economic Bulletinno. 6 / 2021 (September).

Square, maría Pilar. 2023 “ an analysis of the evolution of The hours Worked Per Territory In Spain: Developments and Divert From Recent developments. ” Bank of spain ’ s Economic Bulletin. 2023 /T 1 ( 14 ).https :// doi.org/ 10 . 53479 / 29650 .

Eurostat. 2021 . “ EU Labour Force Survey. Explanatory Notes (to Be Applied from 2021 Q 1 Onwards). ”https :// ec.europa.eu/eurostat/documents/ 1978984 / 6037342 /EU-LFS + Explanatory + notes + from + Q 1 + 2021 + onwards.pdf.
— — —. 2024 a. “ Hours Worked per Week - Comparability across Countries. ” https :// ec.europa.eu/eurostat/databrowser-backend/api/public/ explanatory-notes/get/Info_note_LFSQ_ 20240604 .pdf.
— — —. 2024 b.Improving the Measurement of Employment in National Accounts – Recommendations and Good Practices – 2024 . Luxembourg: Publications Office of the European Union.
— — —. 2024 c.The Application of the Employment Method for the Exhaustiveness of GDP Estimates – Practical Guidelines for Enhanced Comparability between Countries –Edition. Luxembourg: Publications Office of the European Union.

Source, an angel. 2010 . “ Linked Series of wage employment and Regional Labour Income. ” UFAE and IAE Working Papersno. 837 (September).https :// recercat.cat// handle/ 2072 / 87995 .

National statistics institute. 1995 .Encuesta de población activa: evolution of the technical characteristics during the reporting period 1964 - 1976 . Madrid : INE.
— — —. 2005 . “ Questionnaire. Encuesta de población Activa. 2005 . ”https :// www.ine.es/inebaseDYN/epa 30308 /docs/epacues 05 .pdf.
— — —. 2021 “ EPA. 2021 (Methodology 2021 ). Questionnaire. ” https :// www.ine.es/inebaseDYN/epa 30308 /docs/epacues 21 .pdf.
— — —. 2025 . “ Encuesta de población Activa. Microdata. ” .

King of the castle, Pillar. 2024 . “ Understanding Unworked Time in Spain. ” CESifo Working Paperno. 11604 .

Torrejón, Sergio, Enrique Fernandez Macías, Ignacio Gonzalez vázquez, and Ildefonso Marqués Perales. 2024 . “The Working Times They Are A-Changing: Trends in Six EU Countries ( 1992 - 2022 ). ” JRC Publications Repository. 2024 .https :// publications.jrc.ec.europa.eu/repository/handle/JRC 139815 .