| Annex 1 Annex 2 Annex 3 Annex 4 Annex 5 | Terms of referenceSurvey questionnaireSources of data: their respective advantages and disadvantagesOECD workBasis of analyses in Section 3 |
In 1996 the Committee on Employment Policies of the 83rd session of the International Labour Conference requested the International Labour Office to develop an expanded range of indicators of labour market performance, and to assist member States in improving the collection of labour market information and widening the range of labour market indicators.
Much of the response to this request is being taken forward in a project known as Key Indicators of the Labour Market (KILM). The KILM project entails developing, maintaining and disseminating a database of up-to-date and relevant indicators for as wide a range of countries as possible. Guidelines and modules for measuring, collecting and analysing the selected labour market indicators are being prepared. The work is being carried out in close collaboration with the ILO Bureau of Statistics and the field structure, in order to achieve consistency and wide coverage in the data.
An initial list of 18 indicators has been chosen for data collection. These indicators will become known as the ILO Key Labour Market Indicators and will cover underemployment, educational attainment, poverty and income distribution, wages, labour costs and productivity. A new ILO publication (ILO Key Indicators of the Labour Market, to be published later in 1999) will present the KILM database, along with succinct descriptions of the situation in labour markets around the world.
The work described above is known as KILM Activity 1 and by its very nature can only concern itself with the more traditional and well-established statistical indicators. Activity 2 of the project will develop and promote new indicators, capable of revealing new insights in the labour market and the changing nature of employment. Bearing in mind that the 18 Activity 1 indicators are almost entirely 'static' or 'snapshot' measurements at a point in time, it was thought appropriate to begin Activity 2 with a study of the sources and analyses currently used to measure labour movement and flows, in other words labour market dynamics (LMD).
This paper presents the results of the study, which are based partly on an examination of literature on LMD, assembled by various means, and partly on questionnaire responses from national statistical offices. A shorter account of the study will be presented in one of the chapters of ILO Key Indicators of the Labour Market, but the ILO is publishing this near-complete version of the consultant's report in advance, as a response to the fast-growing interest in the measurement of labour market dynamics.
Gek-Boo Ng
Chief
Employment and Labour Market Policies
Branch
Increasingly, analysts and commentators realize that the most familiar and frequently used labour statistics - stock measurements of the number and composition of the employed and unemployed and derived net flows - provide only a limited contribution to understanding the workings of labour markets. This realization is not confined to labour specialists. There is a growing appreciation of the pivotal role of labour markets in understanding economic and social developments generally and a consequential widening of interest in labour dynamics.
This is because a labour market is a system in continual movement. Counts of employment, unemployment and people outside the labour force give only still pictures at a point in time; this is inadequate even when the comparatively rapid snapshots of a monthly series are available. Much of what is going on is not revealed. To shed light on the factors that underlie net changes in stocks, figures are wanted of gross as well as net flows between the count dates or periods; also the labour market history of people and other analytical units over several years.
This paper describes the ways in which the need for data describing labour market dynamics (LMD) are being met; the Terms of Reference are reproduced at Annex l. The study is based partly on an examination of documents compiled in literature searches by ILO Headquarters, within which OECD work - much of it published in Employment Outlook - is given a deservedly high profile; and partly on responses from 40 countries to a questionnaire (see Annex 2). The report also uses material assembled by the author when organizing a one-day session on LMD, held as part of the Conference of European Statisticians (CES) in June 1996, including the organizer's discussion paper prepared for that occasion. The author's attendance at meetings of the Groupe de Paris on Labour and Compensation over the past two years and the July 1998 KILM Workshop have also benefited this study. Collectively, the material gathered from these several sources falls short of being fully comprehensive, but should be generally representative of recent practices and, thereby, should bring out the main issues that need to be addressed in taking the topic further forward.
The layout of this paper is as follows: Section 2 deals with the coverage and other conceptual issues relating to LMD. Section 3 describes the database of references created for this study and provides some analyses derived from it. Section 4 discusses some key aspects that have emerged during the study. Section 5 draws some conclusions and offers some pointers for the future.
There appears to be no definition of the full scope of LMD in common use; in fact there have been scarcely any attempts to define the subject. This probably reflects the fact that, for most labour statisticians, it is not a recognized topic that can or should be separated from other labour statistics. In drawing a boundary round the subject it must be acknowledged, notwithstanding the introductory remarks made above, that the conventional stock figures are not devoid of information on dynamics. Time series of stock data give net changes which tell us whether or not inflows exceed outflows. More pertinently, a single observation of an unemployment stock analysed by short- and long-term unemployment is, in effect, an LMD statistic (and is included in the 18 Activity 1 KILM indicators as No.10). The same can be said about an employment stock analysed by length of job tenure.
Labour Accounting Systems (LAS) are intended to be a comprehensive framework for looking at labour markets and therefore should encompass LMD. They have been described by Eivind Hoffmann as a means for 'the description and analysis of the state and dynamics of the labour market and its interaction with the rest of the economy (author's underlining) ....[including].....studies of 'gross' changes ('flows') in the number of jobs and persons and their activities'. The author provides a description of LMD data in an LAS context as follows:
'...User[s] ...will focus on various changes, such as
(c) the net changes between reference periods in the number of persons in each status category;
(d) the total number of changes occurring in the reference period;
(e) the total number of persons who experience at least one change within the reference period; and/or
(f) the number of persons who have changed status from one period (or one reference date) to the next.
The numbers for (c)-(f) are only equal for short reference periods - periods which are too short for a post or a person to experience more than one change. We must expect that in practice an LAS will mostly be concerned with (c) and (f) type changes for reasons relating to the availability of data ...
... In discussions about LAS a fair amount has been said about 'flows' in the sense of gross changes from one reference date to the next, for example item (e) ... above. Making sure that all possible forms of such changes have been identified and estimated, given the periodicity and reference periods, is one of the 'accounting relationships' necessary within the LAS ...
On ... issues ... such as a core set of change (or 'flow') tables ... it would seem useful to develop international guidelines.'
LAS have scarcely developed beyond the theoretical stage and the few applications in individual countries have so far concentrated largely on stock data. However, this study has shown that Austria, Germany and Switzerland do have LMD data (gross flows) in an LAS framework.
2.3.1 Some examples
To convey the idea of data describing LMD to other people, one has to resort to
near- tautologies such as 'statistics that directly measure movement in labour markets'.
More specific definitions tend to be rather lengthy; for example, a definition presented
at the 1996 CES session was:
Measurement over time of changes in the activity status of individuals (employed, unemployed and inactive) and of changes in jobs of employed persons either in terms of persons who have experienced changes; or the duration of completed spells in status or job situations.
This definition excludes the dynamics of income from employment, and the demography and dynamics of workplaces. The latter is of interest in itself - describing aspects of entrepreneurial job-creating and job-destroying activity - and also as a means of explaining some changes in jobs and activity status.
2.3.2 Sources and methods approach
Another approach - perhaps particularly suitable for an audience of statisticians
- is to get them to think in terms of the sources and methods that are used to generate
data describing LMD. The classification used for the 1996 CES paper was as follows:
1. Surveys of individuals and households
1.1 Recall questions
1.2 Rotation sample designs
1.3 Longitudinal surveys
2. Establishment surveys
3. Registers and administrative records: persons
4. Registers and administrative records: establishments
5. Combination and other methods
This categorization was utilized to demonstrate that there is no ideal single source of LMD data - all of them had relative advantages and disadvantages; these are reproduced from the CES paper at Annex 3.
2.4.1 Classification of LMD analyses
This 'sources and methods' approach was elaborated in the ILO questionnaire
used to collect and code information on current practices for this report, but was also
supplemented by a classification of types of LMD analyses. These are reproduced for
convenience below and are also shown in the questionnaire included in Annex 2 (Questions 3 and 4 respectively):
Sources and methods for data describing LMD
| A
B C D E F G H I J |
Ad hoc survey of persons or households: recall questions
Survey at regular intervals of persons or households, with a different sample each time: recall questions Survey at regular intervals of persons or households: rotational sample design Longitudinal survey of individuals or households: cohort design Longitudinal survey of individuals or households: panel design Survey of workplace establishments: ad hoc Survey of workplace establishments at regular intervals Registers or administrative records: persons Registers or administrative records: workplace establishments Other methods or combination of above sources |
| Type of LMD analysis | |
| A
B C D E F G H I J K L M |
Gross flows measuring transitions between labour market 'states'
Labour and job turnover rates The creation and termination of jobs Births and deaths of firms and their life cycle Job tenure Job security Frequency and length of periods of unemployment Aspects of labour market flexibility and mobility The profile of individuals' earnings over a period of time Absences to have or raise children and re-entry to the labour market The transition from full-time education into the labour force The move from work into retirement Other |
Explanatory notes for most of these 'sources and method' codes were appended to the questionnaire; explanatory notes of 'types of analyses' mainly took the form of illustrations of LMD statistics (see pages 6-8 of the questionnaire included in Annex 2). By these means it is hoped that respondents to the questionnaire shared our understanding of the scope of the topic.
2.4.2 Design of questionnaire
The general approach to the design of the questionnaire was governed by two
factors:
This approach meant that the questionnaire could not be too lengthy nor complicated and thus imposed constraints on the classification systems used for 'sources and methods' and 'type of analysis' questions.
2.4.3 Classification of sources and methods
Regarding the classification systems used for 'sources and methods', although
we have used this familiar term, it is more a classification of 'sources' than 'methods'.
Even for 'sources' there is room for further elaboration. But for 'methods' there is much
highly relevant information which was not gathered, for example on:
If we had attempted to collect all this information it would have resulted in a questionnaire of great complexity which only a few enthusiasts would have completed. However, respondents were asked to supply references for their 'sources and methods' material and this gives us the means to probe further, possibly at some future stage.
2.4.4 Classification of types of analysis
Regarding the classification systems used for 'types of analysis', the approach
was to use reasonably familiar research headings topics that respondents would easily
recognize. A problem with this approach is that the ideal properties of classification
systems of exhaustiveness and mutual exclusion are largely absent. There is work to
be done in future to classify the types of analyses rigorously in this way. As a first step
towards developing such a typology, the 'type of analysis' used in the questionnaire is
mapped against the graphical expression of the familiar 3 X 3 transitions matrix as
shown below. This shows the six flows between the three basic labour market states
of employment (E), unemployment (U) and 'not in the labour force' (N).

| LMD analyses | Relevant flows | |
| A
B C D E F G H I J K L |
gross flows between labour market 'states'
labour and job turnover rates [B(i) (hirings) #] [B (ii) (separations layoffs and quits # ] the creation and termination of jobs* births and deaths of firms and their life cycle* job tenure* job security* frequency and length of periods of unemployment* [G(i) unemployment inflows and outflows #] aspects of labour market flexibility and mobility* the profile of individuals' earnings over time* absences to have or raise children and re-entry to the labour market* the transition from full-time education into the labour force* the move from work into retirement* |
1,2,3,4,5,6
1,2,5,6,7 (2,6,7) (1,5,7) ......... ......... 1,2,5,6,7 1,2,5,6,7 1,2,3,4 1,2,3,4 1,2,3,4,5,6,7 1,2,3,4,5,6,7 3,4,5,6 3,6 1,4,5 |
|
# With the benefit of hindsight, these categories should have been included
explicitly in the questionnaire, as they are among the more common forms of
analysis | ||
This exercise brings out the following points:
(i) short-term gross flows indicating a single change of status; the focus here is on the status and the number of individuals entering and leaving it. Examples are categories A and G(i) above which - besides being statistics in their own right - are also used as supporting explanatory statistics for net changes in stock data.
It is worth commenting further on (c) in the previous paragraph. Although the usual analytical unit in LMD is the individual person, a full understanding also requires data on other analytical units:
Moving from matters of general scope to particular elements of LMD studies, the investigations for this study show that the most sustained attempts to conceptualize or define specific elements of LMD - and (by implication) harmonize concepts and definitions internationally - have emanated from the OECD - either published in Employment Outlook (EO) or associated with the OECD in some other way. A digest of their work over the past ten years or so is presented in Annex 4. Particularly useful features of OECD work are the meticulous recording of national sources of LMD data (a valuable input to this study) and the frankness with which the formidable difficulties of international comparisons in this field are discussed.
The most frequently visited subject in Annex 4 is labour turnover - decomposed into hirings and separations, the latter further decomposed into quits (voluntary separations) and layoffs (involuntary separations). A distinction is made between labour turnover and job turnover, and there is an associated interest in job, enterprise and employer tenure, which appear to be alternative names for the same concept. All their work is valuable to the furtherance of LMD studies, but to the author's mind their special contribution is to draw attention to the need for more and better enterprise-based data, and the contribution this would make to understanding the process of job creation and loss. There is also interest in job-losers and displaced workers and unemployment turnover, i.e. flows in and out of unemployment relative to the stock of unemployed people.
This section presents analyses from the database of references, which has entries for:
Each entry is grouped by country and coded, where possible, according to the type of:
Three working documents were created to support the project:
Annex 5 explains the analysis base for each of the tables in this section and their relationship to each other.
Table 3.l ranks the 80 countries on the database by the total number of useful entries (column d) and each case shows the make-up of those entries as follows:
Column a. Country-specific literature entries identified, received, examined and coded;
Column b. Questionnaire entries received and coded;
Column c. OECD studies received, examined, and coded (see also table at para. 34 in Annex 4);
The table also shows, for the record:
Column e. Country-specific literature entries identified but not coded because they were not received;
Column f. Questionnaire dispatched but not received (also includes 'nil returns' reported indicated by #);
Column g. Total entries on the database.
[N.B. OECD studies (and others by international organizations) identified but not received are excluded from this and other tables.]
It can be seen from column d of Table 3.2 that for 17 countries no information was obtained (Côte d'Ivoire and those below), leaving 63 countries for which information has been used to produce the remaining tables in this section.
For each of these 63 countries, Table 3.1 shows the origin of the information on the database. For example, for Germany, entries were obtained from all three - literature references, questionnaires and OECD studies. In contrast, the only information available for Greece and Portugal was from OECD studies and the only information about Tunisia was from the questionnaire they completed. Mainly due to the effect of column c, the upper part of the table is dominated by OECD countries. Pages 26 and 27 in Section 4 discuss the coverage of developing countries. Overall, OECD studies contributed 38 per cent of the useful row entries, literature 34 per cent and questionnaires 28 per cent.
3.3.1 Sources and methods by country
All 63 countries identified in Table 3.l for which we have useful information are
shown in Table 3.2, indicating the number of times a particular source and method
occurs in the database. The column headings used for the table closely follow those
used in the questionnaire - see paragraph 2.4 above; the meaning of the letter codes
is shown at the end of the table. There is some duplication between the database
references for a particular country - either because different research projects identified
in the literature have used the same source, or because official statistical agencies
have reported the same source in their questionnaire response as the literature or an
OECD study. No attempt has been made to remove the duplication for two reasons: (a)
in nearly all cases the derived analyses were different and had to be preserved to show
complete analyses in Tables 3.3 and 3.4 to 3.10; (b) retaining all references is
necessary to indicate the relative popularity and use of each source. The final row,
beneath the total, shows the number of countries for which at least one column entry
was recorded.
Table 3.1 Country coverage of database entries
| Literature coded |
Q'naires coded |
OECD studies |
Total a+b+c |
Literature not coded |
Q'naires no reply |
Total d+e+f | |
| a | b | c | d | e | f | g | |
| United States* | 38 | 5 | 15 | 58 | 5 | 63 | |
| United Kingdom* | 34 | 15 | 49 | 7 | 56 | ||
| Canada* | 10 | 10 | 12 | 32 | 32 | ||
| Germany* | 9 | 10 | 13 | 32 | 32 | ||
| Australia* | 2 | 14 | 10 | 26 | 26 | ||
| France* | 9 | 14 | 23 | 1 | 24 | ||
| Finland* | 5 | 5 | 9 | 19 | 19 | ||
| Norway* | 4 | 3 | 7 | 14 | 1 | 15 | |
| Netherlands* | 1 | 4 | 9 | 14 | 14 | ||
| Italy* | 2 | 11 | 13 | 1 | 14 | ||
| Sweden* | 2 | 3 | 8 | 13 | 1 | 14 | |
| Denmark* | 2 | 10 | 12 | 1 | 13 | ||
| Austria* | 1 | 4 | 6 | 11 | 11 | ||
| Russian Federation | 4 | 6 | 10 | 3 | 13 | ||
| Japan* | 2 | 8 | 10 | 1 | 11 | ||
| New Zealand* | 2 | 3 | 5 | 10 | 10 | ||
| Spain* | 2 | 6 | 8 | 1 | 1 | 10 | |
| Belgium* | 1 | 7 | 8 | 1 | 9 | ||
| Poland* | 4 | 3 | 1 | 8 | 8 | ||
| Switzerland* | 2 | 3 | 3 | 8 | 8 | ||
| Czech Republic* | 1 | 4 | 1 | 6 | 6 | ||
| Mexico* | 3 | 3 | 6 | 6 | |||
| Argentina | 2 | 3 | 5 | 5 | |||
| Brazil | 2 | 3 | 5 | 5 | |||
| Hong Kong (China) | 5 | 5 | 5 | ||||
| Ireland* | 4 | 4 | 1 | 5 | |||
| South Africa | 1 | 3 | 4 | 1 | 5 | ||
| Greece* | 4 | 4 | 4 | ||||
| Paraguay | 4 | 4 | 4 | ||||
| India | 3 | 3 | 1 | 4 | |||
| Latvia | 3 | 3 | 1 | 4 | |||
| Lithuania | 3 | 3 | 1 | 4 | |||
| China | 3 | 3 | 3 | ||||
| Ethiopia | 3 | 3 | 3 | ||||
| Nicaragua | 3 | 3 | 3 | ||||
| Portugal* | 3 | 3 | 3 | ||||
| Romania | 1 | 2 | 3 | 3 | |||
| Slovenia | 1 | 2 | 3 | 3 | |||
| Tunisia | 3 | 3 | 3 | ||||
| Hungary* | 2 | 2 | 1# | 3 | |||
| Chile | 1 | 1 | 2 | 2 | |||
| Colombia | 1 | 1 | 2 | 2 | |||
| Croatia | 2 | 2 | 2 | ||||
| Egypt | 2 | 2 | 2 | ||||
| Luxembourg* | 2 | 2 | 2 | ||||
| Malaysia | 2 | 2 | 2 | ||||
| Turkey* | 1 | 1 | 2 | 2 | |||
| Bulgaria | 1 | 1 | 1 | 2 | |||
| Estonia | 1 | 1 | 1 | 2 | |||
| Israel | 1 | 1 | 1 | 2 | |||
| Kenya | 1 | 1 | 1 | 2 | |||
| Korea, Rep. of* | 1 | 1 | 1 | 2 | |||
| Peru | 1 | 1 | 1 | 2 | |||
| Philippines | 1 | 1 | 1 | 2 | |||
| Ukraine | 1 | 1 | 1 | 2 | |||
| Venezuela | 1 | 1 | 1 | 2 | |||
| Indonesia | 1 | 1 | 1 | ||||
| Macedonia | 1 | 1 | 1 | ||||
| Morocco | 1 | 1 | 1 | ||||
| Pakistan | 1 | 1 | 1 | ||||
| Slovak Republic | 1 | 1 | 1 | ||||
| Sri Lanka | 1 | 1 | 1 | ||||
| Tanzania | 1 | 1 | 1 | ||||
| Côte d'Ivoire | 0 | 1 | 1 | 2 | |||
| Zimbabwe | 0 | 1 | 1 | 2 | |||
| Algeria | 0 | 1 | 1 | ||||
| Botswana | 0 | 1 | 1 | ||||
| Central African Rep. | 0 | 1 | 1 | ||||
| Djibouti | 0 | 1 | 1 | ||||
| Georgia | 0 | 1 | 1 | ||||
| Lesotho | 0 | 1 | 1 | ||||
| Malawi | 0 | 1 | 1 | ||||
| Mauritania | 0 | 1 | 1 | ||||
| Mauritius | 0 | 1# | 1 | ||||
| Papua New Guinea | 0 | 1 | 1 | ||||
| Swaziland | 0 | 1 | 1 | ||||
| Uruguay | 0 | 1# | 1 | ||||
| Uzbekistan | 0 | 1 | 1 | ||||
| Viet Nam | 0 | 1 | 1 | ||||
| Zambia | 0 | 1 | 1 | ||||
| Total entries | 166 | 133 | 184 | 483 | 31 | 27 | 541 |
| * OECD countries # 'nil' return | |||||||
Table 3.2 shows the contrast between countries such as Australia, some EU members and those in North America - for which LMD studies using all or nearly all of the sources are available; and others such as many developing and transition countries where only one or two sources are available. Looking at the column totals:
Table 3.2 Number of references to each source and method type, by country
| Country | A | B | C | C1 | D | E | F | G | H | I | J |
| Argentina | 2 | 2 | 1 | 1 | |||||||
| Australia | 9 | 7 | 3 | 2 | 4 | 3 | 2 | ||||
| Austria | 1 | 1 | 1 | 1 | 3 | 1 | |||||
| Belgium | 1 | 1 | 1 | 1 | |||||||
| Brazil | 1 | 1 | 1 | ||||||||
| Bulgaria | 1 | ||||||||||
| Canada | 2 | 7 | 4 | 1 | 1 | 6 | 2 | 6 | 6 | 6 | 2 |
| Chile | 1 | 1 | 1 | 1 | |||||||
| China | 1 | 1 | 1 | ||||||||
| Colombia | 1 | 1 | 1 | ||||||||
| Croatia | 1 | 1 | |||||||||
| Czech Republic | 1 | 2 | 1 | 1 | |||||||
| Denmark | 1 | 2 | 4 | 1 | |||||||
| Egypt | 1 | 1 | |||||||||
| Ethiopia | 1 | 1 | 1 | ||||||||
| Estonia | 1 | ||||||||||
| Finland | 1 | 6 | 4 | 1 | 1 | 6 | 4 | 1 | |||
| France | 3 | 4 | 1 | 2 | 2 | 1 | 6 | 1 | |||
| Germany | 1 | 2 | 4 | 7 | 5 | 3 | 5 | ||||
| Greece | 1 | 1 | |||||||||
| Hong Kong (China) | 3 | 1 | 1 | ||||||||
| Hungary | 1 | 1 | |||||||||
| India | 1 | 1 | 1 | ||||||||
| Indonesia | 1 | 1 | |||||||||
| Ireland | 1 | 1 | |||||||||
| Israel | 1 | ||||||||||
| Italy | 1 | 4 | 1 | 6 | 1 | ||||||
| Japan | 2 | 2 | 5 | 1 | |||||||
| Kenya | 1 | 1 | 1 | ||||||||
| Korea, Rep. of | 1 | ||||||||||
| Latvia | 1 | 1 | 1 | ||||||||
| Lithuania | 1 | 1 | 1 | ||||||||
| Luxembourg | 1 | ||||||||||
| Macedonia | |||||||||||
| Malaysia | 1 | 1 | |||||||||
| Mexico | 1 | 2 | 2 | 1 | 1 | ||||||
| Morocco | 1 | 1 | |||||||||
| Netherlands | 3 | 3 | 1 | 3 | 1 | 4 | |||||
| New Zealand | 2 | 2 | 1 | 2 | 2 | ||||||
| Nicaragua | 1 | 1 | 1 | ||||||||
| Norway | 1 | 2 | 2 | 2 | 5 | 3 | 1 | ||||
| Pakistan | 1 | 1 | |||||||||
| Paraguay | 1 | 1 | 1 | ||||||||
| Peru | 1 | 1 | |||||||||
| Philippines | 1 | 1 | |||||||||
| Poland | 1 | 2 | 1 | 1 | 1 | ||||||
| Portugal | 1 | 1 | |||||||||
| Romania | 1 | 1 | |||||||||
| Russian Federation | 2 | 4 | 4 | 4 | 1 | 1 | 1 | ||||
| Slovak Rep. | 1 | ||||||||||
| Slovenia | 1 | 1 | 1 | 1 | |||||||
| South Africa | 1 | 1 | 2 | 3 | |||||||
| Spain | 1 | 3 | 4 | 1 | |||||||
| Sri Lanka | 1 | ||||||||||
| Sweden | 1 | 1 | 1 | 1 | 2 | 2 | 3 | 2 | |||
| Switzerland | 2 | 1 | 1 | 2 | |||||||
| Tanzania | 1 | 1 | |||||||||
| Tunisia | 2 | 1 | |||||||||
| Turkey | 1 | 1 | 1 | 1 | |||||||
| Ukraine | 1 | ||||||||||
| United Kingdom | 1 | 15 | 6 | 1 | 1 | 7 | 7 | 10 | 5 | 5 | 3 |
| United States | 11 | 8 | 20 | 1 | 9 | 3 | 3 | 8 | 1 | 6 | 6 |
| Venezuela | 1 | 1 | |||||||||
| Total | 49 | 81 | 80 | 3 | 15 | 30 | 45 | 79 | 52 | 51 | 43 |
| No. of countries
(out of 63) |
27 | 32 | 31 | 3 | 6 | 10 | 28 | 35 | 24 | 14 | 24 |
| Codes used for column headings in Table 3.2
A Ad hoc survey of persons or households: recall questions B Regular survey of persons or households with a different sample each time: recall questions C Regular survey of individuals or households: rotational sample design C1 As C but using recall questions D Longitudinal survey of individuals or households: cohort design E Longitudinal survey of individuals or households: panel design F Survey of workplace establishments: ad hoc G Survey of workplace establishments at regular intervals H Registers or administrative records: persons I Registers or administrative records: workplace establishments J Combinations of above sources, or other methods | |||||||||||
3.3.2 Type of analysis by country
All the 63 countries identified in Table 3.l for which we have useful information
are shown in Table 3.3, indicating the number of times a particular type of analysis
occurs in the database. The column headings used follow those used in the
questionnaire and the meaning of the letter codes is shown at the end of the table. The
numerical basis of analysis in this table is much greater than in Table 3.2 - and there
are fewer gaps - reflecting the fact that, many times, more than one type of analysis
was reported for a given source (see Annex 8 for further explanation). The final row,
beneath the total, shows the number of countries for which at least one column entry
was recorded.
There were 141 instances of measures of 'Gross flows measuring transitions between labour market 'states' (column A) from 39 countries. The most frequently reported type of analysis was 'labour and job turnover rates' (column B ) - 45 countries. Other analyses frequently recorded were:
C. The creation and termination of jobs - 31 countries;
E. Job tenure - 32 countries;
G. Frequency and length of periods of unemployment - 36 countries.
Table 3.3 Number of references to each analysis type, by country
| Country | A | B | C | D | E | F | G | H | I | J | K | L | M |
| Argentina | 2 | 3 | 3 | 1 | 1 | 2 | 1 | 3 | 2 | 2 | 1 | 2 | |
| Australia | 12 | 8 | 3 | 2 | 9 | 2 | 10 | 6 | 1 | 4 | 1 | 1 | |
| Austria | 2 | 8 | 3 | 3 | 3 | 2 | 5 | 3 | 1 | 2 | |||
| Belgium | 3 | 3 | 3 | 1 | 3 | 1 | 1 | ||||||
| Brazil | 1 | 2 | 2 | 4 | 1 | 1 | |||||||
| Bulgaria | 1 | ||||||||||||
| Canada | 7 | 13 | 9 | 7 | 8 | 5 | 5 | 4 | 5 | 2 | 1 | 1 | 1 |
| Chile | 1 | 1 | 1 | 1 | |||||||||
| China | 3 | 1 | 1 | 1 | 1 | 1 | |||||||
| Colombia | 2 | 1 | 1 | ||||||||||
| Croatia | 1 | 2 | 2 | 1 | 1 | 1 | 2 | ||||||
| Czech Republic | 3 | 1 | 1 | ||||||||||
| Denmark | 4 | 4 | 5 | 3 | 4 | 2 | 1 | 1 | 1 | ||||
| Egypt | 1 | 1 | 1 | 1 | |||||||||
| Ethiopia | 1 | 1 | 2 | 1 | 1 | 1 | 1 | ||||||
| Estonia | 1 | 1 | 1 | ||||||||||
| Finland | 3 | 6 | 4 | 3 | 4 | 2 | 3 | 1 | 1 | 1 | 1 | ||
| France | 7 | 7 | 7 | 5 | 3 | 1 | 4 | 3 | 1 | 3 | |||
| Germany | 10 | 10 | 9 | 5 | 5 | 2 | 4 | 6 | 2 | 1 | 1 | 3 | |
| Greece | 3 | 1 | 1 | 1 | |||||||||
| Hong Kong (China) | 1 | 2 | 1 | 3 | 3 | 2 | 1 | 1 | 1 | 3 | |||
| Hungary | 2 | ||||||||||||
| India | 1 | 1 | 1 | 1 | |||||||||
| Indonesia | 1 | ||||||||||||
| Ireland | 3 | 1 | 1 | 1 | 1 | ||||||||
| Israel | 1 | ||||||||||||
| Italy | 3 | 4 | 7 | 3 | 1 | 1 | 1 | 1 | |||||
| Japan | 2 | 6 | 4 | 3 | 3 | 1 | 1 | ||||||
| Kenya | 1 | 1 | 1 | 1 | 1 | 1 | |||||||
| Korea, Rep. of | |||||||||||||
| Latvia | 1 | 1 | 1 | 1 | 2 | 3 | 1 | 1 | 1 | 1 | 1 | ||
| Lithuania | 1 | 2 | 1 | 1 | 2 | 3 | 2 | 1 | 1 | 1 | |||
| Luxembourg | 1 | 1 | 2 | 1 | |||||||||
| Macedonia | |||||||||||||
| Malaysia | 2 | ||||||||||||
| Mexico | 3 | 2 | 1 | 2 | 1 | 1 | 3 | 1 | 1 | ||||
| Morocco | 1 | 1 | |||||||||||
| Netherlands | 9 | 6 | 4 | 2 | 6 | 2 | 5 | 2 | 2 | 2 | 3 | 3 | 2 |
| New Zealand | 3 | 2 | 4 | 3< |