In addition to categorising occupation, many countries also distinguish between different employment statuses. By this is meant differences in the conditions of employment, rather than whether or not someone is employed. The most common, basic distinction is between those who are employed by another and the self-employed, those who work on their own account. A small number of countries make further distinctions within these categories, but there is considerable variation between them as to which ones are made. There are a number of issues in deciding how far to take account of different categories of employment status in these cases.

As a basic principle, there is a general predisposition in the CAMSIS approach to use as much information about occupations as possible. In the same way that the most detailed occupational classification is used, so also it is desirable to take full account of employment status. This is equivalent to regarding, for example, ‘Self-employed plumber’ as a different ‘occupation’ from ‘Employee plumber’. As with the distinctions between occupational groups, the differences may turn out to be insignificant, but it is valuable to be able to demonstrate this, rather than to assume it. The fact that different countries have different employment status categories creates no problem within the CAMSIS approach because it can deal with different forms of employment status categorisations in the same way that it can deal with different occupational classifications; in the end, it is the scores that matter because they are comparable across countries.

However, there are practical limitations to this approach. The cross-categorisation of several hundred occupational categories by employment status results in many cells with very low numbers. This is particularly the case for those employment statuses, such as employer, that have relatively small frequencies overall. A consequence of the low cell counts is that many of the cross-classified groups are too small to be considered as representative. So, in practice, the level of detail desired is not actually attainable and it becomes necessary to make decisions about aggregating some of the smaller groups or combining them with larger ones. These decisions are made, as far as possible, on the basis of preliminary analyses using different levels of categorisation by employment status, including an occupational title-only version. These analyses may suggest that some employment status distinctions, made in a particular country, are disregarded.

The other important issue is that these fine employment status distinctions may not be made even in all official data and they are often less likely to be made by other researchers, particularly where data are being collected as part of an international comparative project. In practice, therefore, data for a particular country may be presented with varying degrees of detail on employment status. It is clearly desirable to be able to provide CAMSIS scores to cover as many of the variants as possible. We have therefore sought to develop a hierarchical schema that allows researchers to use the level of detail that is available in the data being utilised.

Table 1 shows the relationship between the employment status categorisations used by different countries. The full range of categories available in the data sets used in the construction of the CAMSIS scores are shown by a ‘++’ in the relevant cell. The table does not include those countries that give no information on employment status (Level 0: Australia, Canada, Colombia, Estonia, Finland, France, Ireland, Kenya, Mexico, Vietnam) or those that give information only on self-employed or employee status (Level 1: Hungary, Sweden).

Table 1 presents employment status in terms of three levels of distinction (columns 1, 2 and 3). (There is also a Level 0, where there is no distinction by employment status.) The levels are, as far as possible, hierarchical, such that the lower levels are nested in the higher ones. However, because of the different categorisations, a purely hierarchical arrangement is not possible. Level 3 lists the finest categories that are available. As can be seen, Germany, the UK and the USA provide three different ways of sub-dividing the category of Employee (2 ). Similarly, the USA offers a different way of dividing the self-employed than do Germany, Turkey or the UK. Several countries, but not the UK, distinguish a category of Family assistants, although in the USA they are specifically described as unpaid.

The categorisation of status in employment recommended by the ILO (ICSE-93) distinguishes the four main categories of Employees, Employers, Own-account workers and Contributing family workers. (A fifth category of Members of producer co-operatives is not found in any of our datasets.) Although it is possible to standardise on a single Level 2 category of Employees, the situation with regard to self-employment is more complicated and it is not quite possible to aggregate all systems into a common form for Level 2. Because it does not distinguish a category of Employers, the USA can only be aggregated into category 1.0, Level 1, rather than into categories 1.1 or 1.2 at Level 2. Since some countries make only a basic distinction between employed and self-employed (and because international comparative datasets often adopt this as the standard), it is desirable also to create a wider self-employed category that also includes contributing family workers. Merging groups in this way allows us to create standardised categories (6, 4, 3 and 5) that correspond to those of ICSE-93.

The categories at the different levels for which CAMSIS scores are provided are shown by the numbers in parentheses listed in Table 1 under each country. These categories are usually ones at the finest level of distinction: those cells where there is both a ‘++’ and a number in parentheses for a particular country. However, where preliminary analyses involving different combinations suggest that certain distinctions give a distorted result, for the reasons referred to earlier, an aggregation at the next higher level is used. So, for example, we have not utilised the distinctions between types of Employees made in the German data. Cells where an aggregate category has been used are shown by a ‘+’. The final version of CAMSIS scores are based on the analysis using these basic units.

Once the scores for the basic categories have been determined, those for higher-level, aggregated categories are calculated as the mean of the lower-level categories that make it up, weighted by the number of cases in that lower-level category. So, for example, in the case of the UK, the score for the Level 2 category 2.1, Employee, is the weighted mean score calculated from the Level 3 categories of Employee, Supervisor and Manager. Similarly, for the USA, it is calculated from the scores for Private sector and Public sector categories.

The CAMSIS index file made available for each country includes two 'status in employment' variables, one coded to the usual national system, the other coded to a common, standardised system, as indicated in the fourth column. This should allow researchers to add a CAMSIS score to their occupational data at whatever level of detail for employment status they choose, ranging from none to the finest available detail.

Back to introductory page.
 
 

Last modified 15 August 2002
This document is maintained by Paul Lambert (paul.lambert@stirling.ac.uk)