Institutional and functional statistics differ primarily in what they aim to describe:

• Institutional statistics aim to describe the institution (the actor) – the enterprise, organization, household.
• Functional statistics aim to describe what the institution does (the activity) – what it produces, what production entails, and the value generated.

Institutional statistics

Institutional statistics meet the information needs related to economic conditions concerning the institution (the actor) as a whole and its relationships with other institutions (actors). It aims to provide a general and overarching view of the institutions in the economic market. Institutional statistics describe the economy using variables that can naturally be linked to institutional units as autonomous and decision-making actors in the market. Some examples of such variables include:

• Interest and dividends
• Corporate taxes
• Insurance compensation
• Pension provisions on the balance sheet
• Financial assets and liabilities
• Equity
• Expenditures and investments in IT equipment
• Research and development (R&D) expenditures

The actor in the institutional statistics is the institutional unit, and its delineation is based on ownership relationships and the ability to make independent decisions regarding economic activities. Institutional units are grouped into sectors. The five domestic sectors at the highest aggregated level are:

• Non-financial corporations
• Financial corporations
• General government
• Households
• Non-profit organisations serving households

The sector classification is based on the economic objectives of the units. Broadly speaking, we can distinguish three different economic behaviours: profit maximization (corporations), utility maximization (households), and societal economic optimization (general government). The economic behaviour permeates the activities of the actors, whether they occur within a clearly defined activity, such as electrical installations, or consist of a mix of activities, such as forestry, production of sawn timber, production of pulp and paper, to energy production.

Functional statistics

Functional statistics meet the information needs regarding input, output, and generated value in economic activities, as well as the connections between activities. They describe economic activities and show how the production of goods and services is interlinked in the economy. An economic activity takes place when an actor combines resources such as capital, raw materials, semi-finished products, and labour to produce a good or a service.

Economic activities are classified according to the Swedish Standard Industrial Classification (SNI), which categorizes activities into industries. The goal of industry classification is for the economic activities of each actor to reflect the typical product or service of that industry.

SNI has 5 levels. At the highest aggregated level, there are 21 different codes, and at the finest level, there are 821 codes. Some examples at the highest level include:
A - Agriculture, forestry, and fishing
C - Manufacturing
I - Accommodation and food service activities
K - Financial and insurance activities
R - Arts, entertainment, and recreation

Typical information that one wants to report for the kind-of-activity unit in the functional statistics includes:

• Production value
• Input consumption of goods and services
• Labour costs
• Number of employees and hours worked
• Fixed gross investments
• Operating surplus
• Value added

Institutional vs. functional statistics

Institutional and functional statistics primarily differ in what they aim to describe.

• Institutional statistics aim to describe the institution (the actor) – the enterprise, organization, household.
• Functional statistics aim to describe what the institution does (the activity) – what it produces, what production entails, and the value generated.

When Statistics Sweden tries to describe actors and activities using statistical units such as enterprise or kind-of-activity unit (KAU), for practical reasons, we must adhere to the administrative or legal units we have in our registers in both institutional and functional statistics.

The kind-of-activity unit (KAU) in most cases consists of a single legal unit. The enterprise consists of a collection of legal units within an enterprise group or a single legal unit if the legal unit is not part of an enterprise group. So, even though the theoretical concepts of actor and activity may, in some sense, be of different types (the former is while the latter happens), Statistics Sweden captures them with the same type of legal units, albeit in different constellations. In this perspective, the difference between institutional and functional statistics is more in the content of variables.

Difficulties in comparing statistical values between institutional and functional statistics

Allocating institutional statistics by industry can be problematic. Because this is a requirement in certain contexts, Statistics Sweden must establish a primary industry for enterprises. When reporting enterprises by industry, caution must be exercised when comparing them with statistics based on kind-of-activity units (KAU), which are also reported by industry. This is because some of the kind-of-activity units (KAU) belonging to an enterprise may belong to a different industry than the enterprise’s main industry and, therefore, may not be included in the corresponding statistics based on kind-of-activity units. Conversely, the institutional enterprises may be reported in a different industry than several of their kind-of-activity units. Therefore, it cannot be generally stated that statistical values for the industry of the institutional enterprises are lower or higher than the corresponding aggregated values for their kind-of-activity units since they may have different industries.