That said, there are two ways of analysing the Nationalist-Unionist proportion in Northern Ireland:

1. Analysis of Election results
2. Analysis of social attitude Surveys

It is the opinion of this researcher that the second is more reliable, as the first assumes equal turnouts by the two communities; something which is not the case in Northern Ireland. Nevertheless, both will be studied here for completeness.

Note: The study of election results will be published here shortly.

Analysis of Social Attitude Surveys

Introduction
A potentially more accurate analysis of opinion on the constitutional issue (ie Nationalist or Unionist) is the social attitude survey. Provided they are carried out well, they represent a much more accurate cross-section of the community, and does not depend only on those who choose to vote:

• It has been found that Unionists are less likely to vote than Nationalists, so that a poll could potentially give a better measure of the changing level of Unionism and Nationalism in Northern Ireland than election results.
• It can be demonstrated that a significant number of Catholics who wish to see the retention of ties to the UK still vote for Nationalist candidates at election time [compare pages 35 and 38 of Northern Ireland Social Attitudes Survey 1995-1996, The Fifth Report].

The survey technique will now be used to derive figures for support of Unionist and Nationalist viewpoints in the past decade.

Data Sources
There are two surveys that will be looked at. The first is the "Northern Ireland Social Attitudes Survey 1995-1996". This is a highly respected statistical breakdown of Northern Ireland society and was carried out to a high quality. It is therefore considered reliable. This gives figures for the relevant question for 1989,1990,1991,1993 and 1994. For a more modern figure, the survey carried out by "Market Research Northern Ireland" in May 1998 will be used. It was analysed by the Institute of Irish Studies (Queen's University) and had the support of all the parties in the Northern Ireland peace talks. It is also considered reliable.

The Data
The following table gives the 'preferred long term policy for Northern Ireland' by religion. (Source: Northern Ireland Social Attitudes Survey 1995-1996, The Fifth Report, p.35.)

The Market Research Northern Ireland survey gave the following results:

Do you want Northern Ireland to remain part of the United Kingdom?
YES: Protestants: 97%, Catholics: 33%

Do you want Northern Ireland to become part of the Republic of Ireland?
YES: Protestants: 2%, Catholics: 70%

The reason why the total Catholic figure adds up to103% in this case is uncertain. It is probably due to people who wish to be part of the UK now, but would still like unification at some future point.

Removal of 'Other's
At this point the terms 'Nationalist' and 'Unionist' will be substituted for the relevant survey questions. Additionally, since these figures do not add up to 100%, they will be scaled so that the Protestants and Catholics for each year separately add up to 100%. This removes those who answered the question as 'don't know', 'other' or did not answer. The scaled data will therefore only include people who answered Nationalist or Unionist. The table below gives this scaled percentage data:

Calculating Combined Figures
It is desired to produce a single value for Unionism and Nationalism for Northern Ireland by combining the Protestant and Catholic values. This must be done using the proportion of the population that is Catholic. This figure is given at each census, the last one being 1991 when the figure was put at 38.4%. However, analysis by Compton, P. and Power, J. for the Fair Employment Commission (FEC) has calculated that the figure was more likely 41.5%. Between the1971 and 1991 census, the proportionate Catholic population rose at a rate of   0.3% per annum. The 1991 figure and this fact were then used to calculate the total proportionate Catholic population at the time of each survey.

The combined figures, based on the above figures, are as follows:

This can be represented by the following graph. (The graph for Nationalism can be obtained by flipping this graph over top-to-bottom.)

This graphs shows that those holding Unionist views was growing slowly until 1992-93. At that point (which coincides with the start of the peace process) there was a sudden fall-off in support for Unionist ideas within the Catholic community. There was no such fall-off within the Protestant community. From 1994 to 1998, the figures indicates that support for Unionist views began to climb again.

This graph is very different from that obtained by examing electoral figures. This is probably due mainly to the lower election turnout by the Protestant community and that fact that many Catholics who hold Unionist views still tend to vote for Nationalist parties when it comes to elections.

In summary, the overlaps of the Catholic-Protestant and Unionist-Nationalist affiliations can be summarised by the following diagram. The square represents the Northern Ireland population of 1998. The black line splits the population into Catholic-Protestant while the diagonal line splits the population into Nationalist-Unionist with the correct overlap in each community:

Overlaps of the Catholic-Protestant and Unionist-Nationalist
divisions in Northern Ireland, 1998.