The Journal

The Subjects

The Books





Out For The Count: A critical examination of the DETR (DTLR) bi-annual count of Gypsies/Travellers in England with special reference to Staffordshire

Angela Drakakis-Smith and Keith Mason


Although nationally the Gypsy/Traveller population is small, it is extensive, and its impact at the local level can be great. The 2001 DETR bi-annual census put the number of caravans at 13309 in England, whilst Liegeois (1995) and Kenrick and Clark, 1999) estimate the population in UK to be around 120,000, which includes those who are housed. Unfortunately, for the purpose of resource targeting and policy-making, this latter group is largely ignored (1): in Britain, Gypsies/Travellers are legally defined as 'persons of nomadic habit of life, whatever their race or origin' (the now repealed 1968 Caravan Sites Act definition which survives in English case law). To date, local authorities have only concerned themselves with those Gypsies and Travellers who have lived in caravans or mobile homes on sites. Such sites are divided into authorised and unauthorised. Authorised sites have planning permission and site licences for each caravan together with a full range of facilities such as running water, electricity, etc. Unauthorised/permitted use sites have none of these and can exist from several days to several decades. The DETR January count for 2001 indicated that 328 authorised council sites existed in England. However, the overall DETR count showed that in England, in January 2001, there were 2608 unauthorised caravans, 6160 authorised on council sites (roughly 18/19 caravans per site), 4541 on authorised private sites giving a total of 13309 caravans. This is a decrease of 179 since January 2000, and a possible shortfall of 239/240 sites.

The paper has two objectives. First, to examine the official statistics on Gypsy/Traveller numbers for an indication of geographical patterning. Second, to comment upon the way in which such broad scale analyses can be compromised by the inadequacies of the underlying data.

The term Gypsies/Travellers will be used throughout the paper, which hopefully signals that not all Gypsies are travellers and that not all Travellers are Gypsies. It also points to the fact that 'mobile' groups and caravan dwellers are a very diverse group with diverse needs and are far from homogenous as the blanket term Gypsy or Traveller might imply. Whether or not all Gypsies are of Romani origin is a debate which continues to this day (see Gheorghe, 1997; Reid, 1997; Stewart, 1997; Hancock, 1997).

Problems of Census Enumeration

For purpose of resource apportionment and planning, the British Census supposedly provides the most authoritative accounting of people and housing in the country. However, the problems of under (and over) enumeration are well known (Marsh, 1993). The 1991 Census, whilst including a question on caravans, failed to either tabulate or publish the result, thus effectively excluding Gypsies/Travellers from the count (Drakakis-Smith, 1997) irrespective of whether or not they had been given a census form to complete (and in Staffordshire, many had not). Whilst Gypsies/Travellers may have been included under the rubric of 'non-permanent dwellings' it would be impossible to distinguish them as a group for needs assessment - one of the reasons for conducting the Census. Whilst this omission could be rationalised on the grounds that the Department of Transport, Local Government and the Regions (DTLR) (formerly DETR) conducts its own bi-annual census of Gypsy/Traveller caravans in England, it does not detract from the exclusion of Gypsies/Travellers and the fact that they are separated from majority society by being counted separately - for whatever reason.

The 2001 Census, with its 'Count Me In' hype, rationalises the questions on housing and ethnicity in terms of change and need assessment at the national and local level (2) (S3; S5; Fact Sheet 2; Fact Sheet 4). Whilst these endeavours might partially succeed using the accommodation question 'Do you live in a caravan?' (3) they may be even less successful using the section of questions on ethnicity since Gypsies/Travellers would be absorbed under the 'White/British' and/or 'other' category. Unless Gypsy/Traveller representatives made a special request that Gypsies/Travellers should not be enumerated as a separate group (although they were designated ethnic minority status under the terms of the 1976 Race Relations Act in 1988 (CRE Vs Dutton) and Irish Travellers were recognised as an ethnic group in 2000 (O'Leary and Kiely Vs Punch Retail, The Times, 29 August 2000)) then this could be regarded as an opportunity missed. Both questions regarding accommodation and ethnicity have the potential to render ambiguous responses.

In the absence of census data, then, this paper attempts to work with figures provided by the DETR (now DTLR) in the form of the biannual count of Gypsy/Traveller caravans in England. These data have been available since July 1978 (when public money was used to build council sites - at first via local authority rates and later as 100 per cent Exchequer grant aid) and currently detail the numbers of Gypsy caravans on unauthorised sites and on authorised (public and private) sites in mid-January and mid-July. Because the DTLR count is organised on a geographical basis - all English local district authorities are required to provide a return (4) - the data can be used to provide a current account of the spatial dimensions of the Gypsy/Traveller community (see Map 1, next page). However, we argue that the value of the national data is somewhat compromised by collection procedures at the local level. These can only be revealed by a close examination of the local mechanisms. This is done using Staffordshire as a case study.

Data/Policy Interaction

The DTLR biannual count is closely interwoven with state legislation and planning and local policy - mainly the 1994 CJPO Act, the Town and Country Planning Act 1990 and Structure and Local Plans (5). Such legislation comes invariably with a history, rarely stands alone, and, when used with a heavy hand, can exacerbate the problems it seeks to solve (Okely, 1983). Thus the legislation which affects Gypsies/Travellers and the count should be viewed together, not just because it is the former that gives obvious rise to many of the geographical patterns seen in the DTLR data, but also because the count is itself a direct response to legislative need.

Click HERE for Map 1: Distribution of Gypsy Caravans (January 2000)

The early Caravan Sites and Control of Development Act of 1960 effectively began the nationalisation and control of all caravan sites and attempted to either eliminate substandard sites or upgrade mobile accommodation bringing it into line with the general upgrading and regeneration occurring in the housing arena (6). Given the antipathy towards Gypsies/Travellers by private site owners and the housed community (which exists even today), many Gypsies/Travellers returned to the road, excluded from adequate site provision.

In 1965 the first census for Gypsies/Travellers was undertaken by the Ministry for Housing and Local Government for Wales (MHLG, 1967). The total Gypsy/Traveller population was counted to be around 15000 - subsequently deemed to be a gross under-estimate (Acton, 1974; Adams et al, 1975). The report also stated that only 17 sites in England and Wales were run by local authorities, accommodating 143 families, and with a further 3356 in need of provision. Clearly the 1960 Act wasn't working.

Such findings, together with concerted lobbying by Gypsies and Travellers themselves and supporters, ushered in the stronger 1968 Act, which on the one hand made it a duty for local authorities to provide sites. However, via designation any unauthorised sites over and above the 'adequate' provision made could be removed. (7) The 1968 Act thus began the process of redistribution and spacing out of Gypsies/Travellers. Designation, in effect, began to exclude Gypsies/Travellers from space by legitimising the process of eviction and entrenching an informal quota system. The official bi-annual count began in 1979 after public funding was introduced to create sites (1978).

Prior to the 1994 CJPO Act, which eventually revoked designation, 156 local authorities, i.e. 34 per cent (DETR data), had been excused from making any further provision. In Staffordshire many sites were full when they first opened - accommodating only those already settled in the area on temporary sites - this meant that children growing up had to find a space elsewhere or return to the road.

Given the amount of Guidance in the form of Circulars (1/94, 18/94, 45/94) and the Good Practice Guide 1998 it could be concluded that the 1994 CJPO Act, introduced during the Conservative administration, was construed as a 'knee jerk' reaction (by the Journal of the Police Federation in 1992) (8). It has also been seen as an over-reaction to new recruits into the mobile way of life, namely New Travellers, Eco-Warriors and Road Protesters, some of whom were created by, and were reacting against, Conservative policy (Hetherington, 2000). The Act effectively advocated self-help for Gypsies/Travellers. It revoked any of the positive aspects of the 1968 Caravan Sites Act - the duty of local authorities to provide suitable accommodation and the resources to do so - whilst it strengthened the punitive aspects making unauthorised encampments illegal, punishable by fine or imprisonment. This was a move, not only to clamp down on raves and trespass, but also an attempt by national government to bring legislation for the accommodation of Gypsies/Travellers into line with its ideological switch to self-help and privatisation. It was also a way reducing the number of spontaneous encampments and of excluding Gypsies/Travellers from the Green Belt (9). The Act displayed contradictions, given the difficulties Gypsies/Travellers encountered when attempting to buy their own land for sites or sought planning permission to develop land for this purpose. If equality of opportunity existed it did not apply to Gypsies/Travellers.

Although couched in terms of need, the census was of caravans and not people. The implications of this are that Gypsies/Travellers are objectified and codified as planning, land-use and environment problems rather than as social/human issues. Whilst the DETR Guidance and Circulars may have attempted to infuse some humanity into the legislation, ultimately it is morally easier to justify the removal of caravans - the term unauthorised is easier to attach to vehicles than it is to people. Although the duty to provide has been repealed, the count continues, but now it is couched in planning terms - i.e. to assist planning authorities with planning and the Planning Inspectorate in its deliberations. Although it is still regarded to be an underestimation, and not without inaccuracies it continues to stand as the only longitudinal record of the distribution of Gypsy/Traveller caravans in England, given that many Gypsies/Travellers were excluded from the published 1991 Census material.

Mechanisms of the DLTR Bi-Annual Count (in Staffordshire)

Although prior to the 1994 CJPO Act the DTLR claimed to base national policies for Gypsies/Travellers on the bi-annual count of caravans, critics have long denounced the accuracy of the data. To allay fears and presumably to encourage performance and better practice OPCS was commissioned to conduct a 'thorough review' of the count procedures (OPCS, 1991). Although it highlighted several areas for improvement - the way in which officers identified sites, Gypsies/Travellers and families, and in the way in which the data were recorded - few of the recommendations were incorporated on the grounds that they would be too costly of local authority time and resources. Neither did the DETR wish to jeopardise the local authority good will - given that the data is collected without recourse to national funding.

Whilst Sibley (1984a; 1984b; 1990) and Halfacree (1996) amongst others have criticised the count in general terms, apart from the OPCS Report, the mechanics of the count have not been reviewed. Thus a small survey was undertaken in Staffordshire (as part of a larger project) in 2000. It sought to determine if the count procedures had been influenced by the Report; if officers were adhering to the DETR guidelines; if the count was being carried out in a more systematic way; and were the data accurate, at least for Staffordshire. To this end all the counting officers (9 in all) in Staffordshire were interviewed.


Procedures for the Count

The majority of officers (8) claimed that they were given clear guidelines for the count, and all agreed that the guidelines were in printed form. The majority agreed that the DETR gave clear criteria with regard to who were to be counted. However, in response to the question who decided who was to be counted, 6 officers said they decided, 2 claimed it was the DETR pro forma, and only one said categorically that it was the DETR. When asked how it was decided who to include, 6 said they referred to the DETR guidelines, one claimed that Gypsies/Travellers themselves would tell him, another said that all those camped illegally and living in a caravan, would be counted, and one proffered that he 'had an idea in his head'. All officers said that they had responsibility for counting but in some cases it was the line manager who had ultimate responsibility for overseeing the completed pro-forma (GS1). Some officers were not sure who received the form once it left their domain.

When asked who ensured accuracy of the collected information, 6 said it was impossible to be accurate since people may not be in their vans when they called (whoever was on site was asked for this information). One officer claimed that it was impossible to cover the whole district on one day and officers 'didn't go around looking for Gypsies/ Travellers'. One officer went only to known sites to count and relied on the police and the public to notify him of any unauthorised sites. One officer stated that no check on accuracy was made. When asked if there was any count criteria for inclusion/exclusion, there was no uniform response (multiple responses were allowed):

2 Followed DETR guidelines
1 DETR definitions used
2 Ticked the box on the form
1 Any unauthorised sites were included
1 Looked at site and those on site to decide
2 Asked who they were
1 None

Of the above, one officer volunteered that he could tell by accent, the type of transport used and the way the camp was set out.

When asked how Gypsies/Travellers were identified, similarly, there was no consensus:

2 Asked who they were
2 Identification based on officers' knowledge
3 Identification based on prior knowledge of site
2 Type of site
1 DETR defines who Gypsies/Travellers are

Additional information to these responses were that: Gypsies/Travellers told them who they were; that even if they asked they wouldn't be told; that trying to get accurate information was impossible; that it was 'recognised that there were different groups within what would be classified as Gypsies/Travellers'. The information was recorded variously on bits of paper, in notebooks or on a previous list and the figures were simply amended.

The Count in Practice

All officers said that the count took place on the same day and the day prescribed by the DETR and that no caravans were moved on before the count day - unless the count day coincided with a court Order. Respondents were asked to describe their own procedures for the count - the responses were far from uniform. One set aside a day and police and colleagues were asked beforehand to 'look out for unauthorised sites' as they went about their duties. Another relied upon public complaints and refuse collectors to ring in. If 'no word' was received by lunchtime then it was assumed that no unauthorised encampments existed within the district. Neither felt that they could check everywhere. Three officers visited all known sites on the day including sites where Gypsies/Travellers had been in the past. One monitored the situation throughout the year and visited sites that were in the local authority on the day. Another visited 'leaders' of sites and as they walked the site together the leader would say who was in each van. One Officer visited unauthorised sites only, whilst another 'looked at the files' and added any 'they had' on the day.

To check on thoroughness, officers were asked if they visited all sites and each caravan. In response to site visits 4 said they did, 3 said they did not; one visited known sites only, whilst another 2 visited unauthorised sites where they were known to exist. To the question do you visit each caravan, 6 officers said they did, 3 said they didn't; one added that only unauthorised sites were visited. When asked if all caravans were visited 3 said yes, 5 said no and one said not necessarily. Thus it would appear that not all sites, or all caravans are visited on the day. Officers were also asked if they checked stopping places regularly - 7 said no and 2 said yes. When asked if they identified unauthorised sites regularly, 3 officers said they did, one said he didn't. The remainder were somewhat ambivalent in that 2 checked only if complaints were received, 2 said their council tried to secure sites against the event of unauthorised re-encampments, whilst another claimed that colleagues kept him informed of any unauthorised sites as they went about their other duties. When asked where the information on families came from if they didn't visit all sites and all caravans the response was as follows (multiple responses were allowed):

13 From other families/neighbours on site
3 From site wardens
4 Local authority records
2 Gypsies/Travellers themselves
1 Bailiffs

To the question who is included in the count, 6 officers said that they included New Travellers, 3 did not and one officer added that this group were not required to be included by the DETR. It was also found that 8 included Irish Travellers, one did not respond, 5 said they counted visitors on site, 3 said they did not and one did not respond.

When asked if the accuracy of their count was monitored by anyone, 7 thought not, 2 thought it was, and one thought it was checked by the County. Once the count had been completed, when asked what happened to the information once it left them, 4 thought it was sent to the DETR, 4 thought it was filed away in the local authority's own records and one said it was used when 'talking to other local authorities'.

A Question of Provision

Whilst officers may not have been au fait with what happened to the information once they had completed the DETR GS1 pro-forma, it was deemed useful to know the kind of information local authorities retained from the exercise. This could help them to assess need and/or to inform any policies they might wish to make or adjust for Gypsies/ Travellers. To this end officers were asked what information was kept by their authority on both permanent and unauthorised sites. Regarding permanent sites, information was kept on age (4) and gender (5), with numbers of persons ranking highest (7) and number of school age children ranking second (6). Two authorities kept information on vehicle registration numbers. Only one kept information on employment status. Regarding unauthorised sites, very little information was retained or sought. Two authorities kept information on age and gender, 3 for family size. Two kept information on school age children (an additional officer was uncertain); one kept vehicle registration numbers. Only one officer asked where families were going to and only 2 asked where they were coming from. One officer recorded some names. None asked families on unauthorised sites if they wanted residential accommodation (i.e. if they intended to stay) and only one officer had asked if transit accommodation was required. This tended to support the view expressed by some officers that the local authority 'just did not want to know'. Records, informal or otherwise were kept however, on numbers of caravans on unauthorised sites throughout the year, which appear to remain at the local level

Accuracy in Staffordshire?

Given the above response, the answer must be - hardly! However, at the local level it is possible to detect and correct errors. Thus an attempt was made to assess the accuracy of the figures for Staffordshire using the DTLR count for Staffordshire and Staffordshire County Council and local authority information.

It was found that the Staffordshire total for January 1999 had been recorded on the DETR count as 228. It was 253. In January 1999 East Staffordshire had 23 caravans on an authorised private site which had not been recorded in the return. In July 1999 Cannock recorded 22 caravans on an authorised public site when the site was private. Lichfield regularly records no sites in their returns when it has 2 or more caravans on an authorised site of long standing and at least 2 on an unauthorised site (equally of long standing). The rationale here is that the sites have been there for so long that the residents are not regarded as Gypsies/Travellers by the local authority. In January 2000 the Stafford Borough Council return was 31 caravans on an authorised council site and 32 on an authorised private site. The reality was 11 on the authorised council and 52 on the authorised private site. Whilst this might be regarded as nit-picking, it does show how error is built into the figures and if the error is repeated for each council then the margin of error can escalate to become too great for any kind of diagnostic testing of data. However, the greatest margin of error/omission exists with regard to unauthorised encampments. The count picks up unauthorised sites on two days per year. However, unauthorised sites were reported throughout the year by almost every local authority in Staffordshire.

Thus it would appear that each local authority has its own particular procedure for carrying out the count. Often, 'the way' is dictated by precedent or on the practice of the counting officer, sometimes irrespective of DETR guidelines and criteria. It is often the case that each officer is afforded a great deal of discretion in how the information is collected. The DETR appears to be not unhappy with this scenario. Many of the officers did the count as a matter of course; it appeared not to be rated as an important activity by many. It also appeared that little had changed in terms of procedure and practice at the local level despite the OPCS Report of 1991.


We hope that this paper has demonstrated the way in which policy and practice interact with data, each influencing the other, and the way in which official data on this issue can obfuscate and control reality - in that local authorities tell national government only what it needs to know. The undercounting of caravans on unauthorised sites is a case in point which serves to conceal the need for sites in some local authorities.

As long as Gypsies/Travellers are excluded from mainstream data collection e.g. the Census (and even the biannual count in terms of non-settled unauthorised encampments) they will be disqualified from need assessment exercises. The question needs to be raised: is this deliberate or accidental?


  1. The rationale being that they are entitled to the same rights and privileges as other house dwellers.
  2. To measure the success of equality of opportunity policies, needs assessment at the local and national level with a view to forward planning and targeting national and local resources and future provision.
  3. Although figures here might be skewed given that others - often the retired and elderly or those seeking a cheaper alternative to private, and a better alternative to social, housing - not just Gypsies/Travellers live in caravans and mobile homes.
  4. Some, however, do not, in which case the DTLR 'estimate' the number or simply enter 'No Return Received' (NRR).
  5. The DETR claimed that the bi-annual count demonstrated that the 1994 CJPO Act 'was working' and therefore the government had no need to reform its policies. The data is also used to assist the planning Inspectorate when they deliberate on appeals for sites (Interview 2000).
  6. The 1960 Act gave local authorities the power to open, close and run sites of all types. Many sites were closed as 'unfit for human habitation' (MHLG 1967) - possibly in line with government policies on 'slum' clearance, popular in the 1950s and 1960s.
  7. 15 pitches in the London and county boroughs. Once this number was reached any others could be unceremoniously removed - presumably to other local authorities who made inadequate, if any, provision. The 1968 Act promised resources for the setting up of sites but these did not materialise until 10 years later in 1978.
  8. This is also quoted in Clements and Campbell (1997:61).
  9. The DoE circular 1/94 suggested that theirs had been a 'privileged' position with regard to planning in the past and would be removed, together with any entitlement to live in the Green Belt. Thus Gypsies/Travellers became placed on a par with any other development, but having to overcome obstacles with which house dwellers would not have to contend (Morris 1998).


Acton, T. (1974) Gypsy Politics and Social Change . London: Routledge and Keegan Paul.

Adams, B. et al. (1975) Gypsies and Government Policy in England. London: Heinemann.

Clements, L. and Campbell, S. (1997) 'The Criminal Justice and Public Order Act and its implications for Travellers' in Acton, T. (ed.) Gypsy Politics and Traveller identity. Hatfield: University of Hertfordshire Press.

Drakakis-Smith, A. (1997) Killing Them Softly? An examination of the recent legislative framework which directly affects Gypsies/Travellers (with particular reference to the Criminal Justice and Public Order Act 1994); the way that such legislation has been informed; and evaluative study of the responses by two local authorities in North Staffordshire in terms of policy, practice and service delivery for Gypsies/Travellers. MA Thesis, Ethnic studies, The Department of Sociology, Social Policy and Social Work. Liverpool: Liverpool University.

DTLR. (2000) Bi-Annual Count Figures for 2000. Department of Homelessness and Housing and Gypsy Sites, Department of Transport, Local Government and the Regions, London.

Gheorghe, N. (1997) 'The Social Construction of Romani Identity' in Acton, T. (ed.) Gypsy Politics and Traveller Identity, Hatfield: University of Hertfordshire Press.

Halfacree, K. (1996) 'The distribution of Gypsy caravans in England, 1979-1994'. Geography Vol.8(1):37-46.

Hancock, I. (1997) 'The struggle for control of identity'. Transitions, Vol. (4) 36-44.

Kenrick, D. and Clark, C. (1999) Moving On: the Gypsies and Travellers of Britain. Hatfield: University of Hertfordshire Press.

Liegeois, J-P. (1995) Roma, Gypsies and Travellers. Strasburg: Council for Europe Publications and Documentation Service.

Marsh, C. (1993) 'The validation of census data'. In A. Dale and C.Marsh (eds.) The 1991 Census Users Guide. London: HMSO.

The Ministry of Housing and Local Governement (MHLG). (1967) Gypsies and Other Travellers. London: HMSO.

Morris R. (1998) 'Gypsies and the planning system'. Occasional Discussion Paper 4, Cardiff: TLRU, Cardiff Law School.

Office for National Statistics (2001), Factsheets 2, 4, 9. London: Census Division.

Okely, J. (1983) The Traveller-Gypsies. Cambridge: Cambridge University Press.

OPCS (1991) The Gypsy Count. A Report commissioned by the DETR (S. Green et al).

Reid, W. (1997) 'Scottish Gypsies/Travellers and the Folklorists', in Acton, T. (ed.) Romani Culture and Gypsy Identity. Hatfield: University of Hertfordshire Press.

Sibley, D. (1984a) 'A robust analysis of a minority census: the distribution of travelling people in England'. Environment and Planning A, Vol.16: 1279-1288.

Sibley, D. (1984b) 'Travelling people in England: regional comparisons'. Regional Studies, Vol.19(2): 139-47.

Sibley, D. (1990) 'Urban change and the exclusion of minority groups in British cities'. Geoforum, Vol. 21(4): 483-488.

Stewart, M. (1997) 'The Puzzle of Roma Persistence: Group Identity without a nation', in Acton, T. (ed.) Romani Culture and Gypsy Identity. Hertfordshire: University of Hertfordshire Press.

Angela Drakakis-Smith
Department of Sociology
University of Bristol
12 Woodland Road

Keith Mason
School of Earth Sciences and Geography
Keele University

Home Page Top of page