Statistical Units

A statistical unit is an entity about which information is sought and for which statistics are ultimately compiled — the unit at the basis of statistical aggregates and to which tabulated data refer. Because the economic entities involved in producing, transforming and consuming energy range from small local producers to large multi-activity corporations with very different legal, accounting and operating structures, energy data compilers need to be aware of the different types of statistical unit available to them [IRES, Ch. VI, para. 6.4, PDF pp. 84–85, 2018].

Observation units vs. analytical units

Statistical units fall into two categories:

• Observation units — identifiable legal/organizational or physical units that are able, actually or potentially, to report data about their own activities.
• Analytical units — units created by statisticians, often by splitting or combining observation units, in order to compile more detailed and homogeneous statistics than observation-unit data alone would allow. Analytical units cannot report data about themselves directly, but indirect estimation and imputation methods exist to derive it.

Countries are encouraged to use analytical units as necessary and feasible, since doing so can improve the accuracy of energy statistics where complex economic entities are active in both energy production and other activities [IRES, Ch. VI, para. 6.5, PDF p. 85, 2018] — recommendations tracker row VI/6.5. Data about a statistical unit’s activities can be collected either directly (census or survey) or indirectly (administrative sources) — see Chapter VII on data collection and compilation for detail (not yet ingested) [IRES, Ch. VI, para. 6.5, PDF p. 85, 2018].

The five statistical units

For practical purposes of collecting energy statistics, five statistical units are differentiated: enterprise, establishment, kind-of-activity unit, unit of homogeneous production, and household [IRES, Ch. VI, para. 6.6, PDF p. 85, 2018].

Enterprise

An economic entity in its capacity as a producer of goods and services is considered an enterprise if it is capable, in its own right, of owning assets, incurring liabilities, and engaging in economic activities and transactions with other economic entities. It is an economic transactor with autonomy over financial and investment decisions, and with authority and responsibility for allocating resources to production. An enterprise may be engaged in one or more productive activities at one or more locations [IRES, Ch. VI, para. 6.7, PDF p. 85, 2018].

Establishment

An establishment is an enterprise, or part of an enterprise, situated in a single location, in which only a single productive activity is carried out — or in which the principal productive activity accounts for most of the value added. One or more secondary activities may be carried out, but their magnitude must be small relative to the principal activity; if a secondary activity is as important, or nearly as important, as the principal activity, the unit is better treated as a local unit (an enterprise, or part of one, engaged in productive activities at or from one location) [IRES, Ch. VI, para. 6.8, PDF p. 85, 2018].

For most small and medium-sized businesses, enterprise and establishment are identical. It is recommended that large, multi-activity enterprises spanning different industries be broken up into one or more establishments, provided smaller and more homogeneous units can be identified for which energy data can be meaningfully compiled [IRES, Ch. VI, para. 6.9, PDF p. 85, 2018] — recommendations tracker row VI/6.9.

In energy statistics, the term plant is often used as an equivalent of establishment [IRES, Ch. VI, para. 6.8, footnote 51, PDF p. 85, 2018].

Kind-of-activity unit (KAU)

Any enterprise may carry out many different activities, energy-related or not. To isolate the part of an enterprise relevant to energy statistics, an analytical unit — the kind-of-activity unit (KAU) — may be constructed. A KAU is an enterprise, or part of an enterprise, that engages in only one kind of productive activity, or in which the principal productive activity accounts for most of the value added. There is no restriction on the geographical area over which the activity is carried out; if an enterprise carries out the activity from only one location, the KAU and the establishment coincide [IRES, Ch. VI, para. 6.10, PDF pp. 85–86, 2018].

Unit of homogeneous production (UHP)

For the most complete coverage, compilers may need an even finer split of enterprise activities: the unit of homogeneous production, a production unit in which only a single (non-ancillary) productive activity is carried out. For example, if an enterprise is engaged primarily in non-energy activities but also produces some energy, the compiler may “construct” an energy-producing unit classified under the proper energy-activity category — while maintaining the autoproducer definition of para. 5.45 if the constructed unit is an electricity or heat producer. The sugar industry, which burns bagasse to generate electricity for its own use, is a worked example. Such data usually cannot be collected directly; in practice they are calculated or estimated by transforming establishment- or enterprise-level data under various assumptions [IRES, Ch. VI, para. 6.11, PDF p. 86, 2018].

Household

The scope of energy statistics also includes consumption statistics for the household sector, using a dedicated statistical unit: the household — a group of persons who share the same living accommodation, pool some or all of their income and wealth, and collectively consume certain goods and services, mainly housing and food. Each member should have some claim on the household’s collective resources, and at least some consumption or other economic decisions must be made for the household as a whole (SNA 2008, para. 4.149) [IRES, Ch. VI, para. 6.12, footnote 52, PDF p. 86, 2018]. In some cases a household may also produce energy products for sale or own use [IRES, Ch. VI, para. 6.12, PDF p. 86, 2018].

Illustrative example: a four-company oil corporation

(Unclear-extraction note: Figure 6.1’s tile diagram is linearized by pdftotext to a caption line with no reconstructible layout; the description below is reconstructed solely from the surrounding body text of paras 6.13–6.19, not read directly off the figure — the same caveat pattern used for Figure 5.1 in Chapter V.)

IRES illustrates the five unit types with an imaginary but realistic large corporation, comprising four companies (A–D) engaged in extraction, transportation, refining and retail sale of oil products, spread across eight geographically distinct locations (“tiles”) [IRES, Ch. VI, para. 6.13, PDF p. 86, 2018]:

• Company A — crude oil extraction (ISIC Rev. 4, Group 061), at two locations, tiles (1) and (2) [IRES, Ch. VI, para. 6.14, PDF p. 86, 2018].
• Company B — pipeline transport of the crude oil (ISIC Rev. 4, Group 493), centred at tile (3) [IRES, Ch. VI, para. 6.14, PDF p. 86, 2018].
• Company C — three refineries at tiles (4), (5) and (6); the refinery at tile (6) also carries out a minor secondary activity of electricity generation (ISIC Rev. 4, Group 351), selling small quantities to third parties [IRES, Ch. VI, para. 6.14, PDF pp. 86–87, 2018].
• Company D — retail sale of motor gasoline and diesel (ISIC Rev. 4, Group 473) at gas stations, tiles (7) and (8); the station at tile (8) also carries out a significant secondary activity of retail sale of food, beverages, tobacco and household equipment (ISIC Rev. 4, Group 471) [IRES, Ch. VI, para. 6.15, PDF p. 87, 2018].

Applying each unit definition to the example [IRES, Ch. VI, paras 6.16–6.19, PDF pp. 87–88, 2018]:

• Enterprise — companies A, B, C and D each individually meet the enterprise criteria (autonomy, capacity to incur liabilities), so the example has four enterprises; the individual tiles do not (except tile (3), which is itself Company B’s whole operation) [IRES, Ch. VI, para. 6.16, PDF p. 87, 2018].
• Establishment — tiles (1)–(7) are each a single establishment, being single-location, single-activity units. Tile (8) has a significant secondary activity (other retail); if that activity’s magnitude is close to the primary activity, tile (8) could be split into two establishments [IRES, Ch. VI, para. 6.17, PDF p. 87, 2018].
• KAU — because a KAU depends only on activity, not location, tiles (1) and (2) together form one KAU (crude extraction), tile (3) is its own KAU, and tiles (4), (5) and (6) together form one KAU (refining). Whether tiles (7) and (8) form one KAU or two depends on the significance of tile (8)’s secondary retail activity [IRES, Ch. VI, para. 6.18, PDF pp. 87–88, 2018].
• Unit of homogeneous production — tiles (1) and (2) together form one UHP, but tiles (4), (5) and (6) do not, because tile (6) also does some electricity production; the installation at tile (6) must be conceptually split into a refinery part (joining tiles (4) and (5) as one UHP) and an electricity-generating part (a separate UHP). A similar split applies to tiles (7) and (8) [IRES, Ch. VI, para. 6.19, PDF p. 88, 2018].

This general UHP-splitting approach is not recommended for energy statistics as a whole (see the recommendation below); its use is justified only in specific cases, such as the constructed autoproducer unit at para. 6.11, and is not usually applied to the whole range of units covered by a country’s statistics. Deviations from the establishment-based recommendation should be carefully weighed, since enterprise, establishment, KAU and UHP can each yield different results when data are broken down by economic activity [IRES, Ch. VI, para. 6.20, PDF p. 88, 2018].

Recommended statistical units for energy statistics

For the inquiries covered by IRES, the statistical unit should ideally be the establishment (and, for the household sector, the household). The establishment is recommended because it is the most detailed unit for which the required range of data is normally available, and because grouping data by kind-of-activity, geographical area and size — common analytical needs — is easiest at the establishment level [IRES, Ch. VI, para. 6.21, PDF p. 88, 2018] — recommendations tracker row VI/6.21.

The choice of unit is nonetheless also guided by the purpose of data collection, user needs, and data availability, so the enterprise may also be used as the statistical unit; in practice, for the majority of (especially smaller) units, establishment and enterprise coincide anyway [IRES, Ch. VI, para. 6.22, PDF p. 88, 2018].

Related

• IRES Chapter VI — Statistical Units and Data Items — the chapter digest this page supports
• Statistical Unit Characteristics — the seven characteristics (identification code, location, kind of activity, period of operation, economic-organization type, legal-organization/ownership type, size) collected about the units defined on this page
• International Standard Industrial Classification of All Economic Activities — the classification cited by the illustrative example’s company/activity codes
• System of National Accounts — source of the household definition (SNA 2008, para. 4.149)
• Electricity and Heat Production — the autoproducer definition (para 5.45) explicitly carried over into the unit-of-homogeneous-production definition