UK Statistics

A Guide for Business Users

David Mort

• Publisher: Routledge
• Year: 2019
• eBook ISBN: 9780429674686
• Edition: 1
• Topic: Social Sciences
• Subtopic: Business General
• Index: Social Sciences

First published in 1992, the aim of this volume is to provide an introduction to the range of UK published statistical sources now available to business users. It is not a comprehensive review of UK statistical publishing but a guide to key sources of information in selected subject areas of particular relevance to business users. The coverage, content, methods of collection and limitations of major titles and services in each subject area are described.

Published statistics are important sources of information for business and industry and most statistics either come from official sources, produced by central government, or non-official sources such as trade associations, professional bodies, market research organizations and economic research institutes.

Examples are used to show the strengths and weaknesses of statistical sources and to compare different sources. A bibliography of all the sources mentioned is included at the end of the book; Appendix 1 gives the names and addresses of contacts for further information and Appendix 2 is a list of selected abbreviations and acronyms.

1 UK Published Statistics for Business

---

In 1966, an Estimates Committee report on official statistics made the observation that ‘accurate, timely and relevant statistics are now generally recognised to be an integral part of the equipment with which government, industry and commerce carry out their economic tasks’ (HMSO, 1966). These words are even more relevant in today’s economic and business climate. The complex and often overlapping social, economic and business mechanisms of the 1990s mean that decision making, in both commercial and non-profit making organizations, is becoming more important but also more difficult. One element in good decision making is the availability of reliable information and access to accurate and up-to-date statistical data is an important part of this process.

Many organizations can generate statistical data relevant to their specific needs from within their own establishments and most organizations do so. For example, a company’s sales records, a local authority’s housing records, or a college’s student records can provide useful indicators of the organization’s general activities and performance when the data in the individual records is aggregated. For information on the social, economic and business environment in which the organization operates, it will probably be necessary to turn to external sources of information such as published statistics. In some instances, an organization might carry out its own research or commission an independent agency to carry out research specific to its needs.

With both approaches, original results should be produced which are geared specifically to the requirements of the organization but this original research can be expensive. Published statistics are unlikely to be able to provide the level of detail and uniqueness of original research but they can still offer a useful picture of a particular topic and at a significantly lower price.

The published statistics covered by this book are generally available, usually in hard copy or machine readable format, and they are produced on a regular basis, perhaps monthly, quarterly or annually. This regular supply of data creates a time series and allows the changes over a period of time to be measured. The statistics are produced by a variety of bodies in the UK but they are generally divided into ‘official’ and ‘non-official’ statistics.

‘Official statistics’ is the phrase used to describe all those statistics published by central government departments or a national statistical office acting on behalf of these departments. ‘Non-official’ sources encompasses all those statistics produced by organizations outside central government and, in this book, includes sources such as local authorities, trade associations, professional bodies, academic institutions, employers’ organizations, banks and other financial institutions, private research companies, private publishers and stockbrokers.

Published statistics are still trying to live down Disraeli’s famous quote of ‘lies, damned lies and statistics’. Occasionally, statistics are produced using dubious methodology or are published to deliberately mislead but most statistics are collected and published in good faith.

What is true is that many statistical series have limitations and these limitations may be caused by the way in which the data have been collected, the way the results have been processed, the way the data are published, the way the statistics are used and analysed or a combination of these. An awareness of these limitations is important if published statistics are to be used effectively and specific examples of problems in the areas noted above will be given in various chapters of the book. The remainder of this chapter considers some general points about the nature and characteristics of published statistics.

Currency

Most statistics are inevitably out of date as soon as they are published because it takes time to collect, process and publish the data. In the 1950s, the then Prime Minister, Harold Macmillan, compared using government statistics to looking up train times in last year’s train timetable. They usually can only tell us what has happened in the past, sometimes months or even years ago. Few statistical publications (there are exceptions such as forecasts and real-time services) tell users what they really want to know: what is happening now and what is likely to happen in the future.

Some series are published relatively quickly but then the figures are revised when more information becomes available at a later date.

Methodology

The methodology used to collect data for statistical series usually varies from one series to the next and, in turn, these differences will affect the final results. The methodology may also change over time in a specific time series making it difficult to compare figures in one series from one period with those in another period.

The majority of statistical surveys are based on samples such as a sample of consumers, a sample of companies or a sample of households. It is usually impracticable to survey the total population of people, households or companies in a given area and a sample, providing it is representative, allows information to be obtained relatively efficiently with the results being used to calculate figures for the total population.

The problem is that one sample is unlikely to be exactly the same as another. Samples are devised in different ways, response rates vary, the questions asked may differ and the answers can be obtained using different methods such as postal questionnaires, telephone interviews, face-to-face interviews, diaries and group discussions. Different methodologies mean that two surveys apparently measuring the same thing are likely to come up with slightly different results.

Some official surveys are statutory while others are voluntary and, in some instances, surveys which were originally voluntary have become statutory.

Not all statistical series are based on surveys but some are by-products of administrative systems and procedures. The monthly unemployment statistics, for example, are based on computer records of people claiming unemployment benefits. The monthly product import and export statistics come not from a special survey but from the requirement for traders to declare their imports and exports for tax and duty purposes. The product information supplied by traders to Customs and Excise is used to produce the statistics.

The advantage of using an administrative procedure to generate statistics is that the data are usually cheap and easy to collect. The disadvantages are that the statistics are influenced by any changes in the administrative procedure and the limitations of that administrative system.

The exact scope of a survey is also important with differences between surveys again producing different results. For example, a survey of sales of telephones based only on retail sales of telephones will produce quite different results from a survey based on all manufacturers’ sales or a survey based on deliveries to the trade.

Definitions of what is actually being measured vary from one survey to another. This is particularly true of non-official statistics where few common definitions exist but there are also problems in comparing some official series. For example, the unemployment figures produced by the monthly count of the unemployed and those produced from the separate Labour Force Survey use different definitions of what constitutes an unemployed person.

Geographical coverage can also vary with some statistics relating to the United Kingdom, others covering Great Britain and some just concerned with England and Wales or the individual countries of the UK. Local area statistics might cover a local authority area, a health authority area or the area covered by a local TV or radio station.

Some standardization across various statistical series has been achieved through the use of a number of classification schemes. These schemes attempt to classify such things as economic activities, people, industries, products and occupations into logical categories and, usually, to describe the constituents of each category. Statistics based on these classifications can usually be compared with one another although the classifications need to be regularly updated to take account of social and economic changes.

Classification schemes appropriate to specific subject areas are described in more detail in the relevant chapters of this book but, as an introduction to the schemes and their jargon, a list of the major classifications is given in Table 1.1.

Table 1.1 Selected classification schemes used for statistical purposes

Standard Industrial Classification (SIC)

VAT Trade Classification (VTC)

Standard International Trade Classification (SITC)

Harmonized Commodity Description and Coding System (HS)

Socio-Economic Classification of the Population

Standard Occupational Classification (SOC)

Publishing

Most statistical data is still published in hard copy format. These statistical titles typically consist of pages and pages of tables or graphs with little or no analysis or interpretation. Usually, interpretation is left to the user which can cause its own problems as the following section shows. Most official statistics conform to this pattern, with one or two exceptions, although more non-official sources incorporate analysis and commentary.

Few statistical titles have detailed notes or explanations on how the figures have been collected although official series are usually the exception to this rule. A survey of non-official statistics in 1984 confirmed this, finding that very few sources gave detailed information on how the data had been collected (Siddall, 1984).

Once published in one source, extracts of series are then repackaged and analysed in other sources and, although this circulation of the data is generally useful, it can create problems particularly when trying to track down the exact nature and scope of the data. Many official statistics reappear in sources such as trade association reports, market research reports, on-line sources and bank and stockbroker reports.

Newspapers and trade journals also comment on official statistics and, in turn, they often include summary data from the above sources in articles and features. Figures from articles and features are then mentioned in other reports and articles. In this continual repackaging process, figures are divorced from their original context and any detailed information on the scope and nature of the data is likely to have been lost.

Interpretation

The final word is left with the user of statistics since data require effective analysis and interpretation if the information is to be useful. Any statistical series can usually be interpreted in a number of different ways while specific elements of the series can be excluded or given more weight in the final analysis. A statistical series, however well produced, can be misunderstood. There is no doubt that data can be deliberately manipulated and misinterpreted and particular statistical items excluded or emphasized to support a certain argument or point of view. In other instances, the different interpretations are genuine and result from the nature of time series data and the legitimate emphasis given to particular items in the series.

The two headlines below relate to different analyses of the same series of UK migration data published by OPCS in 1985:

MORE PEOPLE ARE ON THE MOVE (The Times, 29 August 1985, p. 3.)

FEWER PEOPLE ON THE MOVE (British Business, 6 September 1985, p. 450.)

Neither source has interpreted the data wrongly but they have each emphasized different time periods in their analysis. The Times article concentrated on England and Wales and compared migration trends in 1984 with trends in 1980 and 1981 when figures slumped. British Business compared 1984 figures with 1983 figures involving all the regions of the UK.

A more calculated example of different interpretations came in the 1987 general election campaign when the Conservative party and the Labour party both tried to score political points from the same statistics on UK Gross Domestic Product (GDP). The government argued that economic growth had averaged 2.7 per cent a year over the period 1982 to 1986 and this was true. The Labour Party took the series from 1981 to 1986 claiming that growth had averaged only 1.5 per cent a year over the seven-year period and this was also true. The government excluded the poor figures associated with the 1980–81 recession while the opposition included these in their calculations.

In some statistical series, changes in the figures in certain time periods can be caused by seasonal factors. Retail sales, for example, usually go up just before Christmas and unemployment figures can be affected by a surge of people suddenly leaving education. To smooth out these seasonal factors, many series publish seasonally adjusted figures as well as unadjusted figures. When a dual series is on offer, the figures are again open to misinterpretation and a user can deliberately opt for whichever figure suits a particular argument.

All these limitations emphasize that published statistics are not a panacea but neither do they deserve the battering they sometimes receive from their critics. Published statistics cannot provide all the answers but, if used with care, and in conjunction with other information sources, they can be effective supports to the decision-making process.

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Table of Contents
• List of tables
• Introduction
• 1. UK Published Statistics for Business
• 2. UK Published Statistics — publishing organizations and general information sources
• 3. Population Data
• 4. Socio-economic Data
• 5. Economic and Financial Data
• 6. Business Statistics
• 7. Labour Market Statistics
• 8. Market Research Data, Advertising and Media Information
• 9. Statistical Examples
• Appendix 1. UK business statistics — contacts and addresses
• Appendix 2. Selected abbreviations and acronyms
• Bibliography: Statistical Sources
• General bibliography
• Index