As further outlined in "What is manufacturing value added?" MVA refers explicitly to national accounts concepts and intends to measure gross value added of all resident producer units carrying out a manufacturing activity.
In theory, the relevant terms are well defined and for an overview can be referred to as:
value added is defined as output minus intermediate consumption
manufacturing activities can be delineated according to the International Standard Industrial Classification of All Economic Activities (ISIC)
resident producer units cover non-financial as well as financial corporations, households, general government and non-profit institutions serving households
In practice, however, considering the complexity of countries' economies, the path towards gaining the relevant granular information for proper estimation of MVA is very long. The question on how to measure value added for ‘all’ resident producer units within a country's territory is key to the extent of potential deviations in published figures. Basic knowledge about as many resident producer units as possible (ideally stored in a statistical business register) is as well decisive as the chosen practical approaches for data collection, processing and publication.
Resident producer units as listed above are quite inhomogeneous and many national statistical offices established own departments to deal with more homogenous sub-groups. Departments dealing specifically with the public sector usually process information on government-controlled units. Similarly, data on the observation of producers under market conditions following price settings as a result of demand and supply are most commonly dealt with in business statistics or even industrial statistics departments (considering additional synergy effects when units with similar main economic activities are grouped together).
Outside the organizational setting of national accounts, industrial statisticians can focus on far more details and conduct censuses and surveys that fulfil data needs besides those of national accounts, too. One major task in industrial statistics, of course, remains the publication of (multi-)annual structural data including estimates on the value added of manufacturing industries.
Survey schemes in industrial statistics are often historically rooted and subject to many country- specific factors, including the completeness and quality of statistical business registers. Moreover, entities belonging to a survey frame need to be able and willing to derive and provide the necessary statistical information from their (business) records. Legal obligations to report questionnaires can support industrial statisticians in their efforts to collect proper data. Such obligations, however, usually apply to registered entities only. This leads to a possibly incomplete picture published in industrial statistics, as not ‘all’ resident producer units might be adequately covered (e.g. shadow economic activities, but also thinking of household producers that are not registered, more commonly referred to as informal sector of the economy).
Estimates of such missing parts can often be done more appropriately in the more general setting of national accounts, as input data from industrial statistics results can be combined with other available information. Linking with data on exports and imports from trade statistics, for instance, could reveal inconsistencies that ultimately make further adjustments of various kinds necessary. Such adjustments can vary significantly from country to country and over time, but are important in providing stable and reliable time series estimates for key economic aggregates in national accounts publications.
Besides a comprehensive overview, the interested reader finds by far more details and profound insights into the major influencing factors for deviations in the UNIDO ISID working paper "Value added in industrial statistics and national accounts. Equal in theory, different in practice".