Contents:

• The Unemployment Rate: Is the Measurement Working?
• The Labour Force Survey (LFS)
• Turn-out and Turn-off?
• A less-divided nation?
• An upturn for turn-out?
• Where does the 2001 voter turnout performance fit into the overall picture

The Unemployment Rate: Is the Measurement Working?

'Find out the number of unemployed people in Britain' goes the instruction. Easy, eh? They all have to 'sign-on', don't they? Well, yes, people who are unemployed and receiving unemployment benefit have to sign-on, but if you count all those signing-on will you get the answer to the original question?

Well, not necessarily, if the UK experience of the last twenty years is anything to go by.

Let's look at this topic in greater detail to see what problems there have been in verifying the unemployment count.

1980s and 1990s Britain saw more than seventeen changes in the measurement of unemployment. Most of these had the effect of lowering the reported rate. The Government Statistical Service (GSS) found itself defenceless against accusations of 'fiddling' the figures. The Department of Employment's statistics office reported that its staff were targeted with abuse at the time, whenever they appeared in the House of Commons.

Let's leave aside questions about the behaviour of MPs, for now. It's useful to know that there are two main measures of unemployment: the Labour Force Survey; and the Claimant Count. Let's look at how both of these series have been criticised.

[Top]

The Labour Force Survey (LFS)

Based on criteria issued by the International Labour Organisation (ILO), the LFS counts the whole population in either unemployment, employment, or economic inactivity. Economic inactivity means the percentage of the population of working age not engaged in the labour force. It is based on a national survey (sample size 80 000) and produces its results from a three-month moving average.

One of the main outputs of the LFS is the long-term unemployment rate. The variable 'LAPU' is calculated using the ratio of the long-term unemployed, (usually those people who have been out of work for more than one year), to the total number of people unemployed.

To some statisticians this method of counting long-term unemployment violates what they call the 'population at risk' principle. The population at risk of being unemployed for more than one year, they say, is not the current number of unemployed people, but the number of people unemployed a year ago. Only those people who were unemployed a year ago are able to be counted as being long-term unemployed today.

Now this is seen as a statistical error. In technical terms that we can nevertheless understand - if you think hard enough - the error is that the long-term unemployed calculation uses a denominator (the number of unemployed people today) to work out the rate that does not include the numerator (the number of people unemployed a year ago).

Right! That takes a bit of thinking about.
Let's look at the written formula to see what this means:

LAPU = LTU_t / U_t * 100

Where:

LAPU is the ratio of long term unemployment to total unemployment.
LTU_t is the number of people long-term unemployed in year t.
Ut is the total number of people unemployed in year t.

The numerator in the formula is the number that would sit on top of the fraction; the denominator sits underneath the numerator.

To simplify the statisticians' accusation: U_t cannot be used, as we don't know what proportion of today's unemployed were also unemployed a year ago.

The formula that would have produced a true measure of long-term unemployment would be:

TLTU = LTU_t / U_t-1 * 100

Where:

TLTU is the true ratio of long-term unemployment to total unemployment.
LTU_t is the number of people long-term unemployed in year t.
Ut-1 is the total number of people unemployed a year before year t.

Statisticians reported that what LAPU literally tells us is the proportion of people currently unemployed who have been unemployed for more than a year. They say that the reported measurement of long-term unemployment confuses statistics about those who have been unemployed for more than a year, with statistics about what has happened to the general level of unemployment over the past year.

The Claimant Count

The Claimant Count (CC) is an administrative series of data, which provides 100% coverage. The statistics are available for areas defined by postcode and are released monthly.

The CC was believed to be understating unemployment. Much of the blame for this was placed at the door of the then Conservative government who had changed the measurement method consistently over the recent years.

But changes in entitlement to unemployment benefit are only part of the story. It is now clear that throughout the 1980s and 1990s, unemployment benefit claimants were being transferred from the CC to sickness benefits.

Indeed, Employment Offices were chasing the rewards of meeting the performance targets that they had been set. Switching people on unemployment benefit to another form of State help was seen, in itself, as a successful outcome. But, the result was inevitable: by the mid-1990s, the UK had more people of working age 'on the sick' than 'on the dole'.

The policy was, in effect, the longest sick-note in history.

[Top]

Turn-out and Turn-off?

What should we make of the UK population's apparent reluctance to take part in the political process? We're not talking about joining political parties and knocking-on-doors before every election. We're not even talking about donning your Barbour and green 'wellies' and heading off to London to 'save the Countryside'. No. We're on about voting; turning up once every four or five years, at your local town hall, church or school, to place one or two crosses on a piece of paper.

It seems that at every recent opportunity to participate in the democratic process (European, local or General elections), we in the UK just can't be bothered. Every politician seems to believe they've lost touch with the voters. And with no small reason.

The General election on June 7th 2001 saw one of the lowest rates of voter participation since World War 1. If you look at the data, then certainly voter turn-out seems to suggest voter turn-off.

UK voter turn-out data for 1945 - 2001 [20K, Excel file]

First things first, then. You ought to visualise the data. Spend some time studying the table before deciding on the format to use to display the information.

Once you have a visual presentation of the data, you should be able to see that 2001 certainly did see a marked fall-off in the percentage of registered voters who actually turned-out to vote.

Try to make some analysis in the form of a set of statements about the data series you have just charted.

UK General Election Voter Turnout : Registered number of voters 1945-2001

Statements about the data series:

Some possible statements might be:

• Fewer than 4 out of every 10 registered voters actually voted.
• The total number of votes cast was the lowest since 1945.
• The proportion of votes cast to number of registered voters fell below 60% for the first time since before 1945

So, the evidence appears to be compelling: the UK voting population's engagement with the democratic process has plummeted to an all-time low. Quite clearly, something had to be done.

But wait! Let's consider what the possible other explanantions there could be for the low turn-out.

• Disenchantment with politics, especially among 18 to 24 year olds.
• Little choice offered by the main political parties.
• What's the point? It makes no difference anyway.

But we need to verify whether low voter turn-out in 2001 really amounts to a 'crisis in democracy', or if something else is going on. Let's think about the background to the 2001 election to try to find out more.

[Top]

A less-divided nation?

Voters may still hate politicians, but be far happier with each other than in the past. In general, are we in the UK more tolerant than before of others? Are the divisions between social class narrower than at any time? Are we a more egalitarian society, in terms of gender, race, sexuality and creed? If you believe these statements to be true, then perhaps it is a sign of political health that we don't feel the need to turn-out to vote.

If we really felt the need to protect our interests, against other sections of our communities, then we would be far more committed to voting. Examples of this were seen in the 2001 vote:

• In the English midlands' constituency of Wyre Forest, for instance, there was a single-issue campaign centred on the proposed closure of the local hospital. A former health service consultant fought against the sitting Labour MP, and won the seat on a turnout of nearly 70%.
  
• The South-West England constituency of Somerton and Frome had been seized by the Liberal Democrats in 1997, by a majority of 130 votes, from the previous Conservative MP. In 2001, then, this was seen as a pivotal battle for both parties. In the event the LibDem majority increased to over 600 votes, on a turn-out of more than 70%.
  
• In the Northern Ireland constituency of Fermanagh and South Tyrone, Sinn Fein ousted the Ulster Unionist Party MP by a majority of 53 votes. Lack of progress on cementing the peace process in Northern Ireland created political tension and uncertainty. Here the turnout was close to 80%.

The lesson from these examples seems to be that turn-out figures seem to be highest in the most deeply divided parts of the UK.

A less polarised political scene?

We have already seen that voters in the UK today seem to be less 'tribal' in their outlook. Perhaps, also, the political parties themselves are less 'extreme'. The modern election campaign seems to be about winning the 'middle-ground' in British politics. In their eagerness for power, the main parties arguably have drifted closer to each other ideologically, than ever before.

Only fifty years ago, it was quite possible for a Labour voter to believe that if the Conservative party won the General election, the National Health Service would be abandoned, or that the country would be dragged into war with the Soviet bloc. Equally, a Conservative voter could quite conceivably have thought that if Labour won the election, there would be a wholesale nationalisation of British industry and an abolition of private schools.

Under these kinds of perceived threats, voter turn-out in both of the General Elections of the early 1950s was very high. In fact the two elections of 1950 and 1951 saw turn-out levels in excess of 80% - the highest since World War 1.

[Top]

An upturn for turn-out?

Clearly, the General Election of 2001 saw an historically low turn-out, when the results are viewed as a whole. But looking at specific marginal seats, or those in which there was a single, important issue dominating, turn-out seems to be stronger.

The lesson for politicians and society seems to be that while there is certainly widespread dissatisfaction with politicians, especially amongst young people, the 2001 election produced such a low turn-out because it was seen as a 'no-contest'. If the country faced a clear decision between parties holding opposing views and with a great deal at stake, we could expect a far higher percentage of registered voters to turn out.

This was the case in the most recent election to have produced a high overall turn-out, (the 1992 vote), when it was far from clear who (John Major or Neil Kinnock) would become Prime Minister. Perhaps the next campaign to galvanise the voting public will be the one dominated by the decision whether or not to join the Eurozone.

Extension Activity

You will have noticed that the chart on UK Turnout at General Elections holds no information for registered voters in 2001.

However, you ought to be able to calculate this figure from the data that is already there.

Total votes cast = 59.2% of total registered voters

26 365 192 = 59.2% of ?

= 26 365 192 / 59.2 * 100
= 44 535 797

What was the increase in numbers of registered voters over the period 1997-2001

Try to account for the increase in the number of people who were registered to vote over the period 1997 to 2001.

Some ideas for you to ponder:

• Population growth over the period
• Legislation affecting voting age
• New voting methods now permitted
• Any other factors

[Top]

Comments on where the 2001 voter turnout performance fits into the overall picture

Using the ratio of total votes cast to total of registered voters, as an indication of turnout, we can make the following comments:

• The 2001 turnout of 59.2% compares with a mean turnout of 75.2% in General Elections since World War 2
• There is a standard deviation of 5.32%
• Approximately 99% of post-war election turnout results lie within +/- 3 standard deviations of the mean turnout figure
• So 99% of elections over this period produced turnouts between 91.16% and 59.24%
• The 2001 election did indeed produce an exceptionally low turnout

However, this all shows how important it is to look carefully at data to see exactly what it indicates. That is what the process of 'verifying' data is all about.

[Top]