7.148.                  With respect to analysing the monetary aspects of labour, e.g., compensation of employees, labour cost and mixed income (see sects. D.2.3-D.2.5 above), the unit for which data are collected should be the establishment, where possible, rather than the enterprise or firm, to the extent that the accounting system used enables a multi‑establishment enterprise to supply data for each establishment. 

7.149.                  The observation period in comprehensive labour cost surveys should cover the 12 months of the calendar year, whenever possible (if not, the usual accounting year should be used) so as to take account of expenditures that occur only annually or irregularly. 

7.150.                  For each economic activity covered by the national programme of labour cost surveys, it is desirable that data should be collected at intervals not exceeding five years. Until such time as major changes occur in labour cost components, owing to changes in social legislation or other factors, data for the years intervening between two surveys could be estimated wherever suitable data on earnings and other elements of labour cost are available. Special investigations of limited scope during the interim period may provide a satisfactory basis for estimates of certain components of labour cost. 

7.151.                  In addition, countries may conduct special establishment‑based surveys or studies to collect industry‑specific data on wages or compensation of employees, such as the Canadian Tourism Sector Compensation Study. Another possibility may be the use of integrated/combined data sets such as the SNA (see sects. E.1-E.4 above). 

7.152.                  In view of the heterogeneity of the self‑employed population and the complexity of measurement of net income of unincorporated enterprises, the measurement of income related to self‑employment (mixed income) should be phased into national programmes of statistics over an extended period of time. 

7.153.                  The choice of an accounting technique for measuring income related to self‑employment should take into account the circumstances in which the self‑employed operate their business and the measurement objective. Two major techniques can be used: 

(a)            Accruals accounting, which measures the profit earned during the reference period, by taking into account receipts and expenses relevant to that period, irrespective of whether or not they have actually been received or defrayed;

(b)            The cash‑flow technique, which measures actual cash received (including the value of production for own use) and paid out (including the value of production given out free or at reduced prices) during the reference period. 

7.154.                  Two basic observation units are relevant to the measurement of income related to self‑employment, depending on the objective pursued: the job and the individual person. 

7.155.                  For the measurement of the income‑generating capacity of different economic activities, the job is the basic entity on which information is to be collected and analysed. 

7.156.                  When the objective is the analysis of the employment‑related well‑being of the population concerned, the desirable unit is the individual person. The individual is also relevant when analysing the relationship between employment‑related income and educational achievements, seniority in employment, work duration, etc.

7.157.              The collection of data on income related to self‑employment should be based on the regular national statistical programmes, using all available sources, including the SNA. 

7.158.              One such source could be general or specialized household surveys, with individual household members as observation units. 

7.592.              Other sources of data include establishment surveys, administrative records (such as income tax and social security records), surveys of small economic units and population censuses.