URISA (1994), p745-754, copyright Urban and Regional Information Association
Abstract:The findings of a comprehensive telephone survey of geographic information systems (GIS) in all 514 British local authorities undertaken in summer 1993 are compared with those of a similar survey carried out during the first half of 1991. The comparisons highlight the rapid rate at which GIS diffusion is taking place in the British local government sector despite uncertainties about its future. They also draw attention to the changes that have taken place since 1991 with respect to the use of GIS in different types of authority and the range of hardware and software facilities being adopted.
In recent years there has been a growing interest in the diffusion of GIS technologies (see, for example, Masser and Onsrud 1993). This is reflected both in the surveys that have been carried out of GIS take-up in particular user groups in various European countries (see, for example, Campbell and Masser 1992, Craglia 1993, and Kiib 1993) and the efforts that have been made within the European Science Foundation's GISDATA programme to develop a systematic framework for the comparative evaluation of national experiences within Europe (Arnaud et al. 1993).
It must be recognised, however, that while studies such as these are useful in giving an overall indication of the nature and extent of GIS diffusion, there is a lack of time series data on GIS diffusion. This is particularly worrying given that rate at which diffusion is currently taking place and the pace of technological change and innovation.
To overcome these deficiencies regular monitoring of diffusion is required. A key criteria for effective monitoring is compatibility. This requires either a comprehensive re-survey of particular sets of GIS users at regular intervals and/or periodic up-dates of the experience of a carefully selected sample of users. Strategies based on the latter carry with them particular problems in situations of this kind because of the extent to which the criteria used to select the initial sample may change as circumstances change. Consequently the only really effective approach which maintains full compatibility is a comprehensive re-survey. Such an approach not only
[End Page 745]
enables strict compatibility criteria to be met, providing that the same questions are repeated at each re-survey, but it also enables a limited number of new questions reflecting either the institutional or technological developments that have taken place since the first survey to be added to the questionnaire.
For these reasons the findings of a re-survey of GIS in British local government that was undertaken by the authors for the Local Government Management Board between June and September 1993 are of considerable interest as they represent the first systematic attempt to monitor the take-up of GIS in a major user community. The survey replicates the questions asked in an earlier survey which was carried out in spring 1991 (Campbell and Masser 1992). Like this survey, it involved a comprehensive telephone survey of all 514 local authorities in Great Britain. As a result these findings not only provide a comprehensive overview of current practice in British local government but also indicate some of the dramatic changes that have taken place over the two and a half years since the original survey. The main findings of the survey are discussed below under the following headings: adoption, system development and technology. (For a brief overview of British local government, see Campbell and Masser, 1992).
Table 1 gives an overall picture of the state of local authority plans for GIS at the time of the earlier survey in 1991 and the 1993 survey. From this it can be seen that the number of local authorities with GIS has risen from 85 in 1991 to 149 in 1993. In summer 1993 29 per cent of all local authorities in Great Britain had GIS facilities as against 16.5 per cent just over two years earlier. A further 50 authorities had firm plans to acquire GIS facilities within a year. Many of these have already set up working parties and undertaken pilot studies to evaluate GIS. A high proportion of these authorities are already in the process of purchasing their GIS facilities or expect to have them in place within six months. As a result of the large increase in GIS adoption between 1991 and 1993 there has been a corresponding fall in the number of authorities considering GIS from 227 to 139. However, it should be noted that the number of authorities with no plans to introduce GIS actually increased over this period from 158 to 176, or 34 per cent of all authorities. The main reasons given by these authorities are the uncertainties associated with the Government's proposals to reorganise local government and lack of finance. A number of authorities also reported that they had set up working parties to consider GIS and had decided not to proceed with the introduction of GIS facilities as a result of their recommendations.
[End Page 746]
Table 2 shows that there are marked variations in the level of adoption of GIS and/or automated mapping facilities according to the type of authority. The highest level of adoption is at the shire county and Scottish region level. In the case of the former, 43 out of the 47 counties, or 91.5 per cent, already have GIS. In the latter case six out of the nine regions had introduced GIS. In contrast, only 61 out of the 333 shire districts (18.3 per cent) and five out of the 53 Scottish districts (9.4 per cent) had acquired GIS by the time of the survey.
As might be expected, levels of adoption in the metropolitan districts fall midway between those of the shire counties and Scottish regions on the
[End Page 747]
one hand and those of the shire and Scottish districts on the other. Some 34 out of the 69 metropolitan districts (49.3 per cent) already had GIS at the time of the survey.
When the findings of the 1993 survey are compared with those from the 1991 survey it can be seen that the biggest increases in both relative and absolute terms with respect to take-up have occurred at the shire district level. The number of GIS in shire districts has risen from 24 in 1991 to 61 in 1993, whereas the corresponding figures for metropolitan districts and shire counties are 22 to 34 and 32 to 43 respectively. The picture for the Scottish regions and the Scottish districts is similar to that for England and Wales but at a lower level of adoption with only two thirds of the regions and less than 10 per cent of the districts possessing GIS.
The findings of the 1991 survey also showed that there were considerable differences between the north and the south of Great Britain with respect to levels of adoption for all types of authority. More than 20 per cent of all authorities in the south already had GIS in 1991 against only 11 per cent in the northern regions.
From table 3 it can be seen that the ratio between north and south has been reduced from 20:11 to 32:24 since 1991. This is largely due to a very high level of recent adoption in the metropolitan districts of the north and to a lesser extent the shire counties and Scottish regions of the north relative to the south. As a result the south to north ratio has changed from 47:10 to 52:45 in the case of the metropolitan districts and from 71:46 to 90:84 in the case of the shire counties and Scottish regions between 1991 and 1993. Insofar as the shire and Scottish districts are concerned, however, the gap between north and south has actually increased in absolute terms during this period from 9:5 to 21:11.
[insert table 3 -- missing, need to rescan]
[End Page 748]
The previous section dealt with the number of authorities with GIS facilities. To analyse GIS usage it is necessary to consider the number of systems purchased by these authorities and their configuration. At the time of the 1991 survey there were 98 systems in 85 authorities. From table 4 it can be seen that the number of systems has effectively doubled since the time of the last survey and that there are now 195 systems in 149 authorities. The tendency towards a proliferation of systems in the shire counties and Scottish regional authorities that was noted in 1991 has become more pronounced in 1993 with 1.74 systems per shire county and 1.83 systems per Scottish region with GIS. Very few shire or Scottish districts possess more than one GIS facility but there were 40 systems in the 34 metropolitan district authorities.
Over 70% of all authorities have purchased systems since the beginning of 1990. The peak year for local authorities to move into GIS for the first time was 1990, when 35 authorities acquired facilities. Since 1990 the number of authorities adopting GIS has declined slightly. Over three-quarters of the systems have been acquired since the beginning of 1990 and that the peak year for acquisition is 1992. However it should be noted that the figures for 1993 are not complete.
At the time of the 1991 survey just over half the 98 systems in local authorities involved more than one department. By the time of 1993 survey, as table 5 shows, the balance had been reversed in favour of systems at the disposal of a single department with 105 or 53.8% of the 195 systems falling into this category. Departmental systems are particularly common in shire counties and Scottish regions where they account for 59 out of the 86 systems in these authorities. In contrast systems involving more than one department account for three out of five systems in the shire districts and Scottish districts. Only in the metropolitan districts is the distribution of single and multi-departmental systems evenly balanced with 19 single departmental as against 21 multi-departmental systems.
[End Page 749]
When these findings are compared with those from the 1991 survey a number of notable differences as well as similarities can be found. In both the shire counties and Scottish regions and the shire districts and Scottish districts the balance between single department and multi-departmental systems remains more or less the same in the two surveys. If anything there has been a tendency to accentuate the differences between these two types of authority with the proportion of multi-departmental systems in the shire and Scottish districts increasing from 58.6% to 60.9% while the proportion of single departmental systems in the shire counties and Scottish regions has increased from 65.9% to 68.6%.
In the case of the metropolitan districts, however, there has been a massive shift in emphasis away from multi-departmental systems between 1991 and 1993. The proportion of all systems in this category has fallen from 77.3% in 1991 to 52.5% in 1993.
The majority of multi-departmental GIS facilities involve only two or three departments and less than one facility in three involves five or more departments. Facilities involving two or three departments are particularly popular in the shire counties and Scottish regions where they accounted for over 70% of all multi departmental applications. However, even in the shire districts and Scottish districts facilities involving two or three departments accounted for over half the total in this category.
When these findings are compared with those from the 1991 survey it can be seen that the proportion of GIS facilities involving two and three departments has increased from 42% to 56.7% overall while the proportion of facilities involving all departments has fallen from 12% to 3.3%.
The general picture that emerges from this analysis, therefore, is one of an increasing emphasis on departmental applications in the shire counties and Scottish regions and multi-departmental applications in the shire and
[End Page 750]
Scottish district authorities. There is also a growing tendency towards the development of joint facilities involving only two or three departments as against large consortia. This is particularly the case in shire counties and Scottish regions.
Some 404 separate departments are involved in the 195 GIS facilities. A quarter of these departments are responsible for planning or development related activities. Another quarter is accounted for by highways and estates departments. IT and technical services are also well represented in this list as are legal services, parks and recreation and the Chief Executive's department. The other category which accounts for 61 out of the 404 departments involved includes big spending departments such as education (10 cases), housing (9), and social services (5) as well as environmental health (14) and buildings and works (4).
In more than a third of the multi-departmental facilities, planning is the lead department. Only IT/computer services and combined technical services are lead departments in ten or more facilities and in fifteen cases there is no single lead department. Planning is particularly important as the lead department in shire districts and Scottish districts and at the metropolitan district level. At the county level, planning, together with highways and the technical services departments are all lead departments in four or five instances.
Planning departments also account for a third of all single department GIS facilities and over half at the shire district/Scottish district level. Other major users of single department facilities are highways/engineering and emergency services especially in the shire counties and the Scottish regions and combined technical services and estates. When these findings are compared with those from the 1991 survey it can be seen that planning departments have generally strengthened their position relative to other loyal authority departments with respect to both multi and single department GIS facilities. During a period when the number of systems doubled the proportion of planning departments active as lead departments rose from 27% to 35% of all facilities. At the same time the proportion of single department systems based within land use planning sections increased from 23% to 33% of all systems. Apart from emergency services which increased in the shire counties the shares of all the other departments either remained static or actually fell during this period. The biggest losers were highways departments whose share fell from 15% to 6% in terms of lead departments and from 31% to 15% in terms of single departmental facilities. -The proportion of single departmental systems developed by estates departments also fell from 23% to 10% during this period.
At the time of the 1991 survey the market leader for GIS software in British local government was ARC/INFO with 22.4% of all systems. Two and a half years later, after a doubling in the number of systems in local government, ARC/INFO has retained its position with 22.0% of the market (Table 6). ARC/INFO is used mainly by the shire counties, the Scottish regions and the metropolitan districts. The market leader in the shire districts and Scottish districts is Axis with 13.1% of the total number of
[End Page 751]
systems. Other software packages used extensively at the shire and Scottish district level are Alper Records and ARC/INFO. ARC/INFO, Axis and Alper Records are also extensively used by metropolitan districts as is G-GP. Two packages which are widely used at the shire county and Scottish regional level are Wings and X Assist.
Table 6 also highlights the diversity of software currently in use as the others category accounts for nearly a third of all applications. This includes more than thirty software packages such as Small World, Spans, Laser Scan, Atlas GIS and System 9 as well as a few in-house systems.
When the results of the 1993 survey are compared with those from 1991 a number of changes in patterns of software use can be seen. At the time of the 1991 survey there were six packages with more than 5% market share: ARC/INFO, Alper Records, GFIS, Hoskyns G-GP, Axis and MacDonell Douglas GDS. Table 7 shows that there are now seven packages with more than 5% market share. These include all but GFIS from the previous list together with MapInfo, Wings and XAssist.
It should be noted that, in addition, to first time purchases, a number of authorities have switched software between the two surveys. The most commonly cited switch was from GFIS to ARC/INFO.
A key factor related to the fall in market share of GFIS is the decline in the use of mainframe systems for GIS. Mainframe installations now
[End Page 752]
account for only 7.7% of all applications as against 25.5% at the time of the 1991 survey. The dominant hardware platform in all types of authority is the workstation followed by the micro. These two hardware types account for 55.4% and 32.3% of all applications respectively. The shift in emphasis from mainframe to workstation is not only due to the popularity of the latter for new installations: thirteen out of the twenty-five authorities which implemented mainframe GIS in 1991 have subsequently switched to workstations.
The findings of the analysis demonstrate the importance of monitoring the diffusion of GIS in major user communities such as local government and highlight the need for further surveys of this kind to be carried out at regular intervals to keep track of recent development.
The 1993 survey also provides a comprehensive overview of GIS facilities in British local government. When these findings are compared with those from the 1991 survey it can be seen that massive changes have occurred in the two and a half year period that has elapsed since that survey with respect to overall levels of adoption and the types of systems being developed. During this period the number of authorities with GIS has increased by more than 50% whilst the number of systems has doubled. Important changes have also taken place with respect to the level of adoption in different types of authority and also in different parts of the country. The take-up of GIS facilities in terms of the number of authorities has levelled off somewhat since 1990 but the findings of the survey show that the market is still buoyant despite uncertainties about the future of local government and finance.
Over this period the market share of mainframe applications has fallen by over a quarter to less than a tenth and there have also been less dramatic but nonetheless significant changes in the market share of different software packages. GIS has become increasingly associated with planning departments in both single departmental and multi-departmental applications.
Arnaud, A., M. Craglia, I. Masser, F. Salge and H, Scholten (1993) "The research agenda of the European Science Foundation GISDATA programme", International Journal of Geographical Information Systems 7(5), 463-470.
Campbell, H. and I. Masser (1992) "GIS in local government: some findings from Great Britain", International Journal of Geographical Information Systems 6(6), 529-546.
Craglia, M. (1993) "Geographical Information systems in Italian municipalities: a comparative analysis", unpublished doctoral thesis, Department of Town and Regional Planning, University of Sheffield: Sheffield.
[End Page 753]
Kiib, H. (1993) "Diffusion of GIS technology in local government in Denmark", paper presented at ESF GISDATA Specialist Meeting on GIS Diffusion in local government in Europe, 28-30 October 1993, Knutsford, UK.
Masser, I. and H.J. Onsrud (eds.) (1993) "Diffusion and use of geographic information technologies", Kluwer: Dordrecht.
The 1991 survey was carried out as part of a research project funded by the Economic and Social Research Council Geographic Information Handling Programme and the 1993 survey by the Local Government Management Board's, Geographic Information Advisory Group.
[End Page 754]