5. Cross-cutting variables and categories

153. The classifications of status at work and status in employment are complemented by a set of variables and categories that provide more detailed information on characteristics associated with the degree of stability and permanence of a particular work arrangement, and for other characteristics that are not reflected in the status at work categories. They provide definitions and categories for types of arrangement that may be represented in several status categories and are therefore referred to as “cross-cutting variables”. They can be combined in output with relevant status categories to construct output classifications relevant for national purposes.

154. Many of the cross-cutting variables are essential elements of the framework. These variables are required to derive the categories in the ICSE-18, or are essential for the compilation of coherent statistics on work relationships – or both. Some are recommended for use in certain contexts only. The variables described in the draft resolution are therefore divided into three groups: (a) required to compile statistics on the detailed categories in ICSE-18; (b) essential for the compilation of coherent statistics on work relationships; and (c) recommended. 

The following cross-cutting variables and categories are proposed: 

• (a) required for ICSE-18:
  • duration of work agreement;
  • type of employment agreement;
  • contractual hours of work;
  • forms of remuneration;
  • job-dependent social protection coverage
• (b) Essential for statistics on work relationships:
  • duration of employment in the current economic unit;
  • usual hours worked;
  • full-time/part-time status;
  • reason for non-permanence of job;
  • preference or not for a non-permanent job;
  • seasonal workers;
  • type of workplace;
  • domestic workers;
  • home-based workers;
  • multi-party work relationships;
  • job-dependent social protection coverage;
  • paid annual leave;
  • paid sick leave.
• (c) recommended: 
  • number of employees in the economic unit in which the worker is employed;
  • main form of remuneration;
  • reasons for preferring a non-permanent job.

156. Detailed proposals for the cross-cutting variables and categories are provided in the draft resolution. Detailed information on the definition the categories included in the crosscutting variables will be included in the Conceptual framework for statistics on work relationships to be provided as a room document at the 20th ICLS. Further notes on specific cross-cutting variables are provided below. 

Duration of the job or work activity and hours of work

157. Since many of the detailed categories in ICSE-18 include jobs which differ significantly in their capacity to provide ongoing and full employment, statistics classified by status in employment, and particularly the subcategories of employees, should be complemented by information on both the duration of the work arrangement and on hours worked.  

158. Two variables on the duration of the job or work activity are proposed in order to provide a full understanding of the stability of work relationships over time and to assess the extent to which workers without permanent employment relationships have ongoing employment and income security. These are Duration of work agreement and Duration of employment in the current economic unit. The first of these measures the time from the beginning to the end of an employment contract or oral agreement and the second measures the time elapsed since the worker started employment with a particular employer. The latter is needed to assess the stability and continuity of employment in cases where a worker has had a series of contracts with the same economic unit. 

159. The draft resolution provides guidance for each variable on the treatment of recurring contracts including those interrupted by seasonality, on situations where work is mediated through labour hire agencies and on recommended measurement practices. A single set of duration categories is proposed for the presentation of statistics on both variables. 

160. Information on full-time/part-time status, hours usually worked and contractual hours of work is essential to gain a full understanding of the nature of work relationships and for the identification of non-standard employment. The resolution does not propose new standards for these concepts but makes reference to the existing international standards for the measurement of working time.¹ 

Reasons for non-permanent employment

161. Information about the reasons why a person has a non-permanent job, and whether this arrangement is voluntary or involuntary, are important to provide a full understanding of statistics on temporary employment and duration of employment. Conceptually three different dimensions capturing various aspects of reasons for non-permanent employment can be identified. The dimension of voluntary/involuntary, the dimension of why the job is non-permanent and the dimension of why the person has a non-permanent job. 

162. The required variable Reason for non-permanence of job focuses on characteristics of the job and may also be used to allow the identification of both seasonal work and paid apprentices, trainees and interns as well as other characteristics of interest. The resolution proposes a minimum set of categories. Other reasons could be identified separately if relevant in the national context. The variable preference or not for a non-permanent job captures the voluntary element by focusing on whether the person took a non-permanent job because he or she did not want a permanent job.

163. The variable reasons for preferring a non-permanent job is associated with the situation of the person and is not required for the provision of information about the nature of the employment relationship. They are nevertheless of significant interest for both employment and social policy as they may represent barriers to accessing the labour market, especially among women.

Type of employment agreement

164. Type of employment agreement provides information on whether a person has a written contract or an oral agreement. At a minimum, categories for “Written contract”, “Oral agreement” and “No contract” are likely to be required as response categories in household survey questionnaires, since some workers such as casual day labourers may not consider that they have any kind of contract or agreement with an employer. The last two response categories, however, should be aggregated in output. The draft resolution provides information on the uses of this variable and notes that categories for individual and collective agreements may be added if required.

Form of remuneration

165. The two variables related to the basis on which the worker is paid are intended to provide information to understand the nature of the employment relationship but not to provide detailed information on remuneration more generally. The variable “forms of remuneration” should include information on all forms of remuneration received in a particular job as it is required to assist with the identification of the status in employment categories. A separate variable “main form of remuneration” provides additional information and should be collected through an additional question when respondents indicate more than one form of remuneration.

166. In some regions “tips from clients” may be the only type of payment received for certain occupations such as waiter or bag fillers in supermarkets. When this is the case it is likely that the workers concerned would meet the criteria to be classified as dependent contractors. 

Type of workplace

167. The variable type of workplace refers to the type of location where the work is usually performed. The draft resolution proposes a minimum set of categories to identify groups of interest, including home-based workers, domestic workers and workers in multi-party employment relationships. Additional categories may be added to reflect national circumstances specific analytical needs. 

Domestic workers

168. The ILO defines domestic work and domestic workers in the Domestic Workers Convention, 2011 (No. 189), as follows: 

• (a) the term domestic work means work performed in or for a household or households;
• (b) the term domestic worker means any person engaged in domestic work within an employment relationship; and
• (c) a person who performs domestic work only occasionally or sporadically and not on an occupational basis is not a domestic worker. 

169. It is widely recognized that domestic workers are particularly vulnerable to exploitation. They are among the lowest paid of all workers and frequently do not enjoy the same social and legal protections as other workers. An ILO report on domestic workers published in 2013 estimated that the number of domestic workers totalled at least 52.6 million men and women across the world in 2010 and that domestic work accounts for 7.5 per cent of women’s wage employment worldwide.²

170. Statistics are needed to measure not only the total number of domestic workers and their characteristics, but also shifts over time between unpaid domestic work provided by household members and paid domestic work, and between paid domestic work performed by workers employed directly by households and those employed by third-party agencies or for profit. This implies the need for measurement of all domestic work in different forms of work, and for statistics that are not restricted to workers who perform domestic work on a regular basis by employees of households.

171. Since the definitions in Convention No. 189 are not sufficiently precise and have various limitations if used directly for statistical purposes, the draft resolution proposes a general statistical definition of domestic work and a definition of domestic workers for use in statistics on employment. This will allow the provision of comprehensive statistics on the various circumstances in which domestic work is performed, including when it is mediated through agencies or Internet applications, and through independent domestic services providers.

172. According to the draft resolution, workers directly engaged as employees by households to provide services for consumption by household members are counted as domestic workers and defined as domestic employees, irrespective of the nature of the services provided. Workers who are not in a direct employment relationship with households are only counted as domestic workers when the work mainly comprises domestic services. 

173. There are four commonly used approaches to measuring domestic workers in household surveys: industry-based; occupation-based; status in employment-based; and based on the relationship to the head of the household. All of these approaches have limitations and are unlikely to identify domestic workers in all employment statuses. Moreover, the use of a special category for domestic employees in the classification of status in employment is not compatible with the current proposals for subcategories of employees. 

174. The draft resolution proposes a solution for the measurement of all domestic workers. Domestic employees are identified using the industry-based approach,³ which may be derived from both the industry and the institutional sector questions. Self-employed domestic workers can be identified based on occupation and type of workplace. Domestic workers employed by agencies are identified by a combination of industry, type of workplace and occupation. The occupations used for the measurement of domestic workers not employed directly by households should be those commonly held by domestic employees, identified on the basis of empirical analysis of occupation data on domestic employees in the national context. Since in many countries it is likely that the numbers of domestic workers with these arrangements will be very small, it may be necessary to undertake such analysis only periodically.

175. The following categories of domestic workers are specified in the draft resolution:

• (a) domestic employees:
  • (i) live-in domestic employees;
  • (ii) live-out domestic employees; 
• (b) domestic workers employed by service providers;
• (c) domestic service providers employed for profit.

176. Live-in domestic employees are employed to provide services for consumption by household members and live on the same premises as their employers. They may be identified in statistical collections on the basis of information on the economic activity of the employer, their type of workplace and their place of residence. They may also be identified by information on the relationship between each person living in the household dwelling and the household reference person. They do not form part of their employer’s household even though they may be provided with accommodation and meals as remuneration in kind. They should be treated as belonging to separate households from the households in which they are employed.⁴

177. Live-out domestic employees are employees of households who provide services for consumption by household members and do not live on the same premises as their employers. They may be identified in statistical collections on the basis of information on the economic activity of the employer, the type of workplace and the place of residence.

178. Domestic workers employed by service providers are engaged as employees by economic units other than private households to provide domestic services in private households, in a multi-party employment arrangement. They may be identified in statistical collections if their occupation is one of those commonly held by domestic employees, the type of workplace is the client’s residence and their status in employment is employee. 

179. Domestic service providers employed for profit provide domestic services to private households as employers, independent workers without employees or dependent contractors. They may be identified in statistical collections if their occupation is one of the occupations commonly held by domestic employees, the place of work is the client’s residence and their status in employment is employer, own-account worker or dependent contractor.

Multi-party work relationships

180. A multi-party work relationship is a contractual arrangement involving a worker, an economic unit for which the work is performed, and a third party. Workers with these arrangements are not directly employed by the economic unit to which they provide their services. They may be employees of the third-party economic unit, or they may be dependent contractors whose access to raw materials, clients or the market is controlled by a third party. 

Employees with multi-party work arrangements

181. The concept of multi-party employment arrangements, whereby a worker is paid by one enterprise to perform work for another enterprise, exists in many countries. Traditionally this group has been referred to as “employees in triangular employment arrangements”, but more recently this has been adapted to “multi-party employment arrangements”. 

182. Multi-party arrangements for employees are mediated by an institutional unit that acts as the employer and makes the worker available, on a temporary or permanent basis, to work for another economic unit while paying the wage or salary of the employee. 

183. Two main groups of employees with multi-party employment arrangements can be identified: 

• (a) Labour hire employees. 
• (b) Employees providing outsourced services.

In some countries, a third group may also be identified:

• (c) Employees in employment promotion schemes.⁵

184. Distinguishing between these three groups is important to inform policy debate. While there is strong demand for statistical information about all groups, the provision of information about the first group (labour outsourcing) should be a priority. Although some workers with these arrangements have very similar social protections and conditions to other workers and may benefit from a degree of flexibility associated with temporary agency work, there are also policy concerns about excessive use of such arrangements and their impact on working conditions.

185. The draft resolution provides definitions and explanatory notes for each of the three subgroups of employees with multi-party employment arrangements. These notes seek to clarify the boundaries between each group, as well as with employees who are not in multiparty arrangements. The blurring of these boundaries, both conceptually and in the minds of survey respondents, however, inevitably results in measurement difficulties. In all of these cases, the work is not mainly performed at the premises of the agency that pays the employee. The type of workplace is usually the premises of the client but may be some other place. This group of workers may therefore possibly be identified in statistical sources through questions on who is responsible for payment of wages and salaries, and the type of workplace. Since the concept of multi-party employment relationships is broad and complex, often changing depending on national laws, it is not possible, however, to provide advice on a single approach to data collection or a single data source. 

186. Household surveys, establishment surveys and administrative registers are all important potential sources of data on multi-party employment arrangements. Several countries have used estimates from administrative sources and/or establishment surveys in preference to household survey data. Household surveys are nevertheless an important and valuable source of data both for the purposes of international comparability, and to obtain complete information on the characteristics of workers with these arrangements. 

187. In view of national differences in the legal situation with respect to labour outsourcing, the optimum measurement approach in household surveys will vary from country to country. When data on multi-party employment arrangements are collected using Labour Force Surveys, two main approaches can be identified: (1) type of employment contract; and (2) who pays the wage or salary. 

188. The first approach involves asking a direct question about the type of employment contract or arrangement that the worker has. This may be suitable in situations where there is a clearly defined and well understood type of employment contract for labour outsourcing. The information may be collected as part of a general question on employment type, or as a direct question such as “Do you have a contract with a temporary employment agency?” 

189. The second approach involves asking a question on whether the worker is paid directly by the economic unit for which the work is performed, or by another agency. This may need to be used in conjunction with a question on type of workplace, in order to exclude those workers who perform work directly for an employment agency, for example, as administrative staff. The distinction between labour outsourcing and subcontracted service provision could be made either by providing a separate response category as part of the “who pays” question (for example, temporary employment agency, labour broker) or based on the industry of the employer. 

190. Several potential problems would need to be addressed at the national level, however, if this second approach is to be used effectively. Firstly, experience has shown that survey respondents may have difficulty in accurately responding to questions on who pays for the work performed. This may be related to identifying who the employer is or to the respondent simply not knowing. 

191. Secondly, correct reporting and coding of the industry of the employer may also be a problem when there are multi-party employment arrangements. In Labour Force Surveys in several countries, the industry of the economic unit for which the work is performed is recorded for agency workers, rather than the industry of the employer. This leads to inconsistency with information collected from establishments and administrative sources and complicates the identification of agency workers. Ideally, the industry of both the employing enterprise and the user enterprise should be recorded in order to properly account for the employment impacts of the user enterprise, identify the extent and nature of labour outsourcing, and provide coherent data for national accounts purposes. This may be difficult to achieve in practice, however, given the complexity of industry coding in household surveys and the difficulties experienced in some circumstances in identifying the employer. 

Dependent contractors with multi-party work arrangements

192. In the case of dependent contractors, multi-party work arrangements exist when an intermediary supplies raw materials and receives the goods produced by the dependent contractors, or else access to clients or work is controlled by an intermediary, typically using the Internet. The contractor may be paid directly by the client, or payment may be received only through the intermediary that benefits from the work performed by the contractor. 

Variables related to the measurement of informal employment

193. The resolution defines three variables related to the measurement of informal employment:

• (a) job-dependent social protection;
• (b) access to paid annual leave; and
• (c) access to paid sick leave.

194. These variables are related to the economic risk to which the employee is exposed and are useful to assess the impact of new or non-standard forms of employment on access to leave and social protection. They are recognized as possible criteria for measuring informal employment among employees in the 17th ICLS guidelines concerning a statistical definition of informal employment. They are also frequently used by countries when measuring informal employment among employees. These three cross-cutting variables are not sufficient to measure informal employment, however. Comprehensive measurement of informal employment also requires the identification of informal employment among workers with all statuses in employment. This would require additional information that is not specified in the draft resolution.

1. ^ Resolution concerning the measurement of working time, adopted by the 18th International
   Conference of Labour Statisticians (November–December 2008), http://www.ilo.org/global/statistics-and-databases/standards-and-guidelines/resolutions-adopted-by-internationalconferences-of-labour-statisticians/WCMS_112455/lang--en/index.htm.
2. ^ ILO, 2013e.
3. ^ For most operational measurement purposes, data on employment classified at only one-digit level to ISIC Rev.4, Section T, Activities of households as employers; undifferentiated goods and services-producing activities of households would provide sufficiently accurate information.
4. ^ SNA, 2008.
5. ^ Persons who are required to perform work in an employment promotion scheme as a condition of continued receipt of a government social benefit such as unemployment insurance are excluded from employment according to the 19th ICLS resolution I (para. 31(c)). They may not therefore be counted as employees but may nevertheless be considered to be in a multi-party work relationship.