Home > Statistics > International trade in services > 2004 > Statistics on international trade in services

Statistics on international trade in services

Methodological description

Target population

The statistical unit in the statistics on international trade in services is an enterprise and the frame population comprises all enterprises in Statistics Finland's Register of Enterprises and Establishments. Apart from data on the population, information is used on enterprises that in the previous inquiry of 2003 reported international trade activities in services and/or are included in the Foreign Affiliates Statistics (FATS) for 2003.

To form the target population of the inquiry, the following were removed from the enterprise population:

1. Closed or inactive enterprises and, basing on legal form, e.g. natural persons and death estates.
2. Enterprises operating in industries unlikely to practise international trade in services and industries which had reported "No foreign trade in services" in the inquiries of the two previous years.
3. Manufacturing enterprises employing fewer than 20 persons or with a turnover of under EUR 10 million.
4. Enterprises in service industries employing fewer than 10 persons and with a turnover of under EUR 1.5 million.
5. Enterprises having reported in the previous inquiry on international trade in services that they do not conduct international trade in services.

The obtained size of the survey frame was 8,422 enterprises which, on the basis of their size and/or industry, could be assumed to practise international trade in services. The survey frame was divided into three sub-frames: 1) total survey enterprises, 2) small and medium-sized enterprises (SMEs), and 3) small enterprises (primarily enterprises employing fewer than 20 persons). The sub-frame of the total survey included enterprises that had reported international trade in services in 2003, employed at least 100 persons or were in foreign ownership (or owned foreign affiliates) and employed at least 80 persons. Using these criteria, the group of the total survey comprised 2,300 enterprises, all of which were thus included in the inquiry.

The sample surveys of the sub-frames 2) and 3 were divided into strata on the basis of 2-digit level of industry and size category of the enterprise. The industry "Other business activities", comprising, among others, diverse technical services and advertising, was divided into more detailed strata using the 3-digit level of the industrial classification in addition to the enterprise size category. Random sampling was made by Bankier's allocation, which allows heterogeneity inside strata to be taken into account. If a stratum contained larger deviation of turnover than usual, more enterprises were taken from it into the final sample. At least five manufacturing enterprises and ten service enterprises were included from each stratum. With the stratified random sampling the final sample covered 13 per cent of the sub-frame of SMEs and 9.5. per cent of the sub-frame of small enterprises.

The size of the inquiry population comprised 2,982 enterprises to which a questionnaire was sent in January 2005.

Definition of the sampling frame and sample in 2004

FATS = Statistics on foreign affiliates
ITS = Statistics on international trade in sevices
SMEs =small and medium-sized enterprises

 

Inflation

After checking the data, the data concerning the sample material were inflated to cover the whole survey frame. For the non-respondent enterprises that had provided information on international trade in services in the previous year, new values were imputed by calculating service type-specific changes in the size of items from 2003 to 2004 and then by multiplying the previous year's responses by the rate obtained. For the other non-respondent enterprises in the sample export and import data were imputed using the ratio of personnel number to exports (or imports) in each industry as the inflation variable. The Horvitz-Thompson model was used for calculating total values for exports and imports by industry. The obtained total values were divided by service type and country using as the ratio variable the ratios derived from the responses of the total survey enterprises, which had been processed so as to eliminate excessive effects on the distribution by clearly deviating observations. The inflation ratio used for the actual total survey was one. This resulted in total statistics concerning international trade in services by service type and target country.

Correctness of the data

By the used statistical method, the aim is to produce a reliable picture of Finnish enterprises' foreign activities in services. The largest sub-frame, i.e. the total survey frame, included all Finnish enterprises that had had international trade in services or which could be shown by other indicators to most probably have had foreign activities of some scale.

In all, 93 per cent of the enterprises responded to the survey. The response rate is excellent, partly due to the statutory nature of the survey. Of the enterprises, 31 per cent reported that they did not have international trade in services. In the previous year, the proportion of such enterprises was larger and the decrease is partly due to the small changes made to the sampling with which the target population could be defined more accurately. In the total survey frame, only 17 per cent of the respondents reported that they had no international trade in services. As expected, the numbers of enterprises not practising international trade in services grew clearly when moving from the total survey to the sample survey enterprises in sub-frames 2 and 3, but in them, too, the numbers of enterprises practising foreign trade in services went up from the previous year.

Response rates and numbers and proportions of 'no' responses by sub-frame in 2000-2004

2000	Enterprises	Respondents	NO1	Response	No / resp. %
Total survey	1,549	1,437	428	93	30
SMEs (2	607	503	339	83	67
Small enterprises	401	295	237	74	80
Total	2,557	2,235	1,004	87	45
2001
Total survey	1,681	1,599	436	95	27
SMEs	625	493	372	79	75
Small enterprises	412	324	276	79	85
Total	2,718	2,416	1,084	89	45
2002
Total survey	1,887	1,805	405	96	22
SMEs	484	415	290	86	70
Small enterprises	325	285	209	88	73
Total	2,696	2,505	904	93	36
2003
Total survey	1,925	1,876	327	97	17
SMEs	496	425	249	86	59
Small enterprises	337	295	204	88	69
Total	2,758	2,596	780	94	30
2004
Total survey	2,300	2,180	443	95	20
SMEs	371	318	204	86	64
Small enterprises	311	275	207	88	75
Total	2,982	2,773	854	93	29

¹⁾ NO =no international trade in services
²⁾ SMEs =small and medium-sized enterprises

For both exports and imports the data of the total survey covered around 92 per cent of the final total data on services. Imputed data made up under one per cent and inflated data seven per cent of total data on exports. The share of imputed data was under two per cent and that of inflated data six per cent of total data on imports of services. In recent years, the significance of imputed data has, on the whole, been lower than in the initial years, because less data have been imputed especially for large enterprises. The improved targeting of the sample has been reflected in an improved response rate and in extended reporting of both exports and imports in the sample survey.

Inquiry surveys always involve possible sources of errors, mainly arising from non-response and over and under-coverage. For the inquiry on international trade in services the response rate is excellent that the impact of non-response inevitably remained small. The coverage of an inquiry survey refers to the correctness of the survey frame in relation to the groups of enterprises the inquiry intends to cover, in this case the group of enterprises engaged in international trade in services. Under-coverage thus means that the survey frame is considered too small and over-coverage that the survey frame includes units not belonging to those to be covered in the survey. Under-coverage and over-coverage can be caused by delays in the updating of various registers, for example in connection with company start-ups, closures or reorganisations. In this survey under and over-coverage did not produce any significant problems because the majority of the respondent enterprises reported international trade in services.

A possible major source of errors in the inquiry on international trade in services is the questionnaire where the problems could arise from the used definitions and the identification of target countries, which are not necessarily analogous with the bookkeeping methods generally used by enterprises. This produces additional work for respondents and raises the possibility for erroneously recorded responses. The effect of possible sources of errors is taken into consideration when developing the inquiry further.

---

Last updated 21.4.2006