We said in our second report that several years would be needed to assess the full effects of the National Minimum Wage. We now have almost two years’ data which allow us to make a more definitive assessment of the initial impact of the minimum wage. We remain of the view, however, that the National Minimum Wage is at a relatively early stage in its development and that more time is needed before a fuller assessment can be made.
Our task this time round is to continue to evaluate the range of direct and indirect effects of the introduction of the minimum wage and of the upratings. We have given special consideration to groups and sectors which have a large proportion of low-paid employees and on which the effects are likely to be greater.
It is not straightforward to isolate the effects of the minimum wage from other changes that affect pay and the labour market, such as changes in the economic cycle, other tax and benefit reforms, and working time and other regulations. A key to our evaluation strategy has been to establish what would have happened in the absence of the minimum wage. We have done this by analysing time-series data before and after the minimum wage was introduced, and by considering groups or sectors operating above minimum wage levels which we used as a control for those who are most affected.
At the whole-economy level, we considered a range of factors covering both the introduction of the minimum wage and its uprating. These include numbers and types of workers benefiting from the minimum wage; the scale of pay increases and the impact upon differentials; the effect on competitiveness, including firms’ total pay bill costs and moves towards more productive working practices; employment, unemployment, and participation in the labour market; and the effect on public sector finances.
As well as covering the whole-economy effects, we monitored those sectors or groups that are most affected by the minimum wage. Among the key low-paying sectors are hospitality, hairdressing, retail, cleaning, textiles and care homes. Among the groups of workers who are likely to be disproportionately affected, and therefore required specific monitoring, are young people, women, ethnic groups, disabled people and homeworkers. The second volume of our report will contain more detailed analysis of these groups and sectors.
Our second report identified several important indirect effects of the minimum wage whereby employers and employees adopt strategies to adjust. Some of these changes occurred in the short run or were anticipated before the minimum wage was introduced; others have longer lead-in times and have only shown up in the current round of research. These dynamic effects may have positive or negative consequences on the economy as a whole, and being able to identify them and assess their importance will be key to determining the continuing success of the new legislation.
A crucial element in our evaluation strategy is the reliability of the statistical data provided by the ONS on the numbers in low pay. We are grateful to ONS for the steps it has taken to develop its methodology, but there remains scope for further improvements. We would encourage ONS to continue to develop its approach to low pay in order that the figures can be regarded as authoritative. We supplemented evidence gathered from official data sources with a research programme, a full consultation exercise and a survey of firms in low-paying sectors.
In our previous reports we highlighted the shortcomings of the official earnings data which had subsequently caused us to revise downwards our estimate of the numbers who had benefited from the minimum wage. At the time of our second report, ONS was working on refining the methodology used to provide estimates of numbers in low pay. As a result, ONS published a ‘revised’ central estimate of the incidence of low pay (ONS Press Release, 26 October 2000). In Appendix 1 we make an assessment of this revised methodology and its estimate of the extent of low pay in the economy.
In summary, our estimate in 1999 that around 1.9 million workers would benefit from the minimum wage was revised downwards in our second report to between 1.5 and 1.7 million workers. Using the refined ONS methodology, we have again revised the numbers downwards. Table 3.1 summarises the numbers and percentages earning below £3.60 (age 22 and over) and £3.00 (18—21 year olds) per hour in 1998, 1999 and 2000 on this revised basis.
| Table 3.1 | ||
| Number of Jobs Paid Below the National Minimum Wage, 1998—2000 | ||
| Year | Number of jobs paid less than specified amounts | |
| (%) | Number (000s) | |
| 1998 (<£3.60/£3.00) | ||
|
18—21 |
7.7 | 120 |
|
22 and over |
6.4 | 1,400 |
|
Total |
6.4 | 1,520 |
| 1999 (<£3.60/£3.00) | ||
|
18—21 |
2.7 | 40 |
|
22 and over |
2.4 | 540 |
|
Total |
2.4 | 580 |
| 2000 (<£3.60/£3.00) | ||
|
18—21 |
3.1 | 50 |
|
22 and over |
1.1 | 250 |
|
Total |
1.2 | 300 |
| Source: ONS | ||
The figures for 1998, in Table 3.1, are not consistent with the 1998 estimates in the second report. In order to allow for wage growth between 1998 and 1999, we based the estimates of potential beneficiaries in our second report on the numbers of jobs paying less than £3.50 per hour (£2.90 for 18—21 year olds). Using the refined ONS methodology, we now estimate that around 1.3 million jobs were paid at these levels in 1998, a significant reduction from our original estimate.
Using the refined methodology, we analysed the groups and sectors that were most affected by the introduction of the minimum wage. Although the overall numbers were smaller, Figure 3.1 shows that the composition was similar to that in our second report. Over 70 per cent of beneficiaries were women, and around two-thirds of jobs below minimum wage levels were part-time.
Figure 3.1
Proportion of Jobs Paying Less than £3.50 (22 and over), £2.90 (18—21) by Gender and Hours Worked, 1998
Source: LPC calculations based on grossed NES and LFS data
As can be seen from Figure 3.2, we found considerable variation between regions, with the North East, Wales, the South West and Northern Ireland having the highest proportions of jobs below minimum wage rates in 1998.
Figure 3.2
Percentage of Jobs Paying Less than £3.50 (22 and over), £2.90 (18—21) by Region, 1998
Source: LPC calculations based on grossed NES and LFS data
Note: Government Office Regions shown for England.
Figure 3.3 shows that particular sectors were also disproportionately affected, with 20 per cent of jobs in the hotel and restaurant sector estimated to be below minimum wage rates in 1998, compared with an estimated 5.3 per cent overall. Wholesale, retail and the motor trade, community social and personal services, and health and social work all had more than 6 per cent of jobs paid below minimum wage levels in 1998.
Figure 3.3
Percentage of Jobs Paying Less than £3.50 (22 and over), £2.90 (18—21) by Industry Sector, 1998
Source: LPC calculations based on grossed NES and LFS data
Note: Sample size too small for reliable estimates in other industry sectors.
Figure 3.4 shows that over 11 per cent of firms with fewer than ten employees were affected by the minimum wage. This was more than twice the national average.
Figure 3.4
Percentage of Jobs Paying Less than £3.50 (22 and over), £2.90 (18—21) by Business Size, 1998
Source: Grossed NES data, April 1998
Our analysis of the LFS suggests that ethnic minority employees were over-represented in the group paid below minimum wage rates in 1998. The sample numbers are small but Pakistani-Bangladeshi employees appear to be over-represented and so stood to benefit considerably from the introduction of the National Minimum Wage.
Low pay is common among homeworkers: those who work mainly at home. Research published last year, looking at LFS data for Spring 1997 to Autumn 1998, found that a quarter of homeworkers were paid below £3.60 per hour and, of these, nine out of ten were women (Felstead and Jewson, 2000). Manual homeworkers were particularly affected: about three-quarters of them were paid below £3.60 per hour. And pay rates among manual homeworkers were on average around half of those of manual workers who did not work at home (£2.86 compared with £5.49). Ethnic minority manual homeworkers were among the lowest paid. Many homeworkers should, therefore, have benefited as a result of the National Minimum Wage.
Disabled workers — those with a work-limiting disability — are generally more likely to be lower paid than non-disabled workers. LFS data show that in 1998 the lowest decile of hourly earnings among disabled workers was £3.62 compared with £3.78 among non-disabled workers. Few of those submitting evidence commented on the impact of the National Minimum Wage on disabled workers’ pay. But we were encouraged by the views of the Advisory Committee for Disabled People in Employment and Training which told how disabled people had benefited from the minimum wage.
As we mentioned in our second report, research by Incomes Data Services Ltd. (IDS) suggested that some employers had been making adjustments to their pay structures in anticipation of the minimum wage. By April 1999 the numbers earning below the minimum wage had fallen to 580,000. And at April 2000 the numbers had fallen further, with an estimated 300,000 jobs paid below National Minimum Wage levels.
|
The introduction of the National Minimum Wage has brought to light examples of workplaces where it had not been previously known that a disabled person doing the same work as others was receiving less pay, just because they were disabled. This practice has had to stop and has been an important positive step forward. Advisory Committee for Disabled People in Employment and Training evidence |
We shall be covering compliance in the second volume of our report. In the meantime, we emphasise that the numbers shown below the minimum wage in the official data cannot be taken as a measure of non-compliance. Neither the LFS nor the New Earnings Survey (NES) monitors compliance with the National Minimum Wage. For example, they cannot identify those such as apprentices, those undergoing training and those eligible for the accommodation deductor, who may receive less than the National Minimum Wage. And the data show some ‘bunching’ just below minimum wage rates which suggests that there are minor inaccuracies in reported hours or earnings, resulting in some jobs being shown as below the minimum wage. Finally, as we said in our second report, similar surveys in countries with long-established minimum wages where compliance is regarded as universal, such as the United States and France, show large numbers of workers below the minimum rates.
In summary, our judgment is that workers in around 1.3 million jobs were entitled to higher pay as a result of the introduction of the minimum wage. Women, part-time workers and people from ethnic minority backgrounds, in particular, stood to benefit. Although the numbers who benefited were lower than we estimated in our first and second reports, the minimum wage has made a significant difference to many low-paid workers.
Earnings
Levels
Figures 3.5 and 3.6 show the change in the hourly earnings distribution, between April 1998 and April 2000, based on the ONS central estimate methodology. The concentration of jobs in April 1999 that paid £3.60—£3.70 per hour, (for both those aged 22 and over and those aged 18—21) had diminished by April 2000, indicating that wages had increased over and above minimum wage levels. At April 2000, around 850,000 jobs for those aged 22 and over paid below £3.70 per hour, and 100,000 jobs for those aged 18—21 paid less than £3.20 per hour, showing that fewer people stood to gain from the June and October 2000 upratings of the minimum wage than from its introduction. Increases in earnings at the bottom end of the distribution have therefore outstripped the increase in the minimum wage.
Figure 3.5
Hourly Earnings Distribution for Those Aged 18—21, April 1998—April 2000
Source: ONS Central estimate methodology
Figure 3.6
Hourly Earnings Distribution for Those Aged 22 and Over, April 1998—April 2000
Source: ONS Central estimate methodology
|
We have seen the emergence of a ‘mezzanine’ floor of around £4.00 an hour as leading firms take steps to keep their rates ahead of the field, driven in part by the tightening labour market. Incomes Data Services, 2000 |
As can be seen from Figure 3.7, in 2000 there were peaks in the earnings distribution for full-time employees aged 22 and over at around £4—£4.10 and £5—£5.10 per hour. The April 2000 NES data show that a greater number of full-time adult workers were paid £4.00 per hour than were paid £3.60 per hour, with a ‘valley’ between the spikes at £3.60 and £4.00 per hour.
Figure 3.7
Hourly Earnings Distribution for Full-time Employees Aged 22 and Over, 2000
Source: Grossed NES April 2000
Table 3.2 shows increases in the bottom decile, median, and upper decile earnings between 1998/99 and 1999/2000. Between these dates the lowest decile earnings increased faster than the median, and earnings growth in the bottom decile outstripped increases in the top decile in 1998/99 and 1999/2000.
| Table 3.2 | ||
| Increase in Earnings 1998—2000 | ||
| Year | Increase in Earnings (%) | |
| 1998/99 | 1999/2000 | |
| Full-time age 18 and over Lowest Decile | 5.4 | 3.3 |
|
Full-time age 18 and over Median |
4.8 | 2.8 |
|
Full-time age 18 and over Upper Decile |
5.5 | 2.9 |
|
Part-time age 18 and over Lowest Decile |
9.1 | 2.8 |
| Part-time age 18 and over Median | 4.8 | 1.4 |
|
Part-time age 18 and over Upper Decile |
5.7 | 2.2 |
| Source: Grossed NES data, April 1998, 1999, 2000 | ||
The differences in earnings between different groups, sectors and regions changed between 1999 and 2000. Women, ethnic minority and disabled workers are all disproportionately represented among the low paid. The minimum wage would therefore be expected to have a positive effect on equal pay, and on the earnings gap of other disadvantaged groups. In our second report we observed that between 1998 and 1999 the gender pay gap (the ratio of female to male hourly earnings) narrowed by a full percentage point. This gap narrowed by a further percentage point between April 1999 and April 2000: average full-time hourly earnings for women increased from 81 to 82 per cent of average male hourly earnings. A more significant increase, over the same period, was seen for part-time workers: female part-time earnings increased from 81 to 89 per cent of male earnings. Not all of this can be attributed to the impact of the minimum wage. Other factors such as changes in the composition of the workforce and changes in hours are also important. A recent study (S. Dex et al., 2000) modelled the effect of the introduction of the minimum wage on the gender pay gap. It found the minimum wage produced small increases in the overall female/male hourly pay ratio, with larger changes for manual workers, and for part-time female employees compared with all men.
As shown in Figure 3.8, there was some narrowing of the relativities in average earnings between sectors over the period 1998 to 2000. Lowest decile earnings across the low-paying sectors identified in Figure 3.8, increased about twice as fast as those across all sectors, nearly 11 per cent compared with 5.3 per cent overall in 1998/99. Since 1999, increases in lowest decile earnings have been more subdued. The largest increases were in agriculture (6.7 per cent) and clothing and footwear (5.8 per cent). Other sectors such as hospitality and cleaning saw very small changes of around 1.5 per cent.
Figure 3.8
Increase in Lowest Decile Earnings for Employees Aged 18 and Over in Low-paying Sectors, 1998—2000
Source: Grossed NES data, April 1998, 1999, 2000
The minimum wage has also had an impact on variations in earnings between regions. Figure 3.9 shows the increase in lowest decile earnings by region between 1998 and 2000. The increase in lowest decile earnings across the UK as a whole was 3.2 per cent in the year after the minimum wage was introduced. The highest increases were in the East, Wales, the South East and the North East. Of these, the North East and Wales are low-paying regions and the higher than average increases there are likely to be a minimum wage effect. But the South East and Eastern regions are higher-paying, and large increases there are more likely to indicate a tight labour market. Other regions appear not to fit into any simple pattern. The evidence suggests that the minimum wage may have narrowed the regional pay gap, but the picture is complicated by the tight labour market and other factors.
Figure 3.9
Increase in Lowest Decile Earnings for Employees Aged 18 and Over by Region, 1998—2000
Source: Grossed NES data, April 1998, 1999, 2000
Note: Government Office Regions shown for England
In our second report, using the NES matched sample between April 1998 and April 1999, we produced estimates of the increase in wages that resulted from the introduction of the minimum wage. We estimated that those paid below minimum wage rates in 1998 and who remained in the sample in 1999, saw an increase in their wages of between 20 and 31 per cent. Not all of this increase was attributable to the minimum wage; some of the increase would have arisen through incremental pay drift or annual pay increases.
It is not possible to revisit the NES figures using the new, grossed NES data because the NES has been grossed for each year separately; the data which link individuals between years, have not been grossed. We would not recommend using the LFS data because, as is explained in Appendix 1, the LFS hourly rate variable changed between April 1998 and April 1999, and therefore comparisons between these years will not be valid.
We asked Mark Stewart of Warwick University to provide some analysis of increases in hourly earnings from the British Household Panel Survey (BHPS). The advantage of the BHPS is the panel nature of the data: the same individuals are tracked over a long time period. In addition, in contrast to the NES, coverage of the earnings distribution is complete. The main disadvantage compared with the NES and LFS is the smaller sample size.
Table 3.3 shows median hourly earnings in 1998 and 1999 for those aged 22 and over, who remained in the same job in the two years, by whether or not they earned less than £3.60 per hour in 1998. It shows that those earning less than the minimum wage in 1998 saw an increase of 15.5 per cent between 1998 and 1999, compared with a 4.6 per cent increase for those earning above minimum wage levels. While none of these estimates is entirely reliable, these data provide further indication of the boost to earnings provided to people in low pay from the minimum wage.
| Table 3.3 | ||
| Increase in Earnings due to Introduction of the National Minimum Wage | ||
| All aged 22 and over, who remained in the same job | Median earnings
in October 1998 |
Median
earnings in October 1999 |
| Earning less than £3.60 in 1998 | £3.16 | £3.65 |
|
Earning £3.60 or more in 1999 |
£7.36 | £7.70 |
| Source: BHPS data, 1998,1999 | ||
In our first report we concluded that ‘if the National Minimum Wage is introduced at a sensible level, the pressure to restore pay differentials following its introduction will be limited and localised’. This conclusion was important in making our initial recommendations because we were conscious of the need to recommend a rate that did not encourage extensive reassertion of pay differentials.
In our second report we noted that the evidence from the NES and research we had commissioned tended to bear this out. Our analysis of the NES suggested that the effect on differentials appeared to peter out at about £5.00 per hour. And IDS research and other surveys suggested that differentials were not being affected greatly or that effects were being dealt with in a straightforward way. Our conclusion was that the restoration of differentials seemed to have been limited.
We received mixed evidence on differentials in preparing this third report. Some organisations reported substantial effects and others reported minimal impact. In its evidence to us, the CBI indicated that ‘of the companies reporting that the NMW had an impact on their business in the CBI Employment Trends Survey 2000, only 11% reported that there had been an impact on differentials’.
This view was supported by the British Retail Consortium (BRC), which argued that because the minimum wage was ‘set at a workable £3.60 (now £3.70), the maintenance of differentials has not been a substantial problem’. The British Hospitality Association stated that ‘we accept that the impact of the NMW on differentials has been limited, but would again stress that, if the factors behind this — a moderate introductory rate of the NMW, and labour shortages — were to disappear, there could be a more significant impact’.
But we received reports of more significant effects, especially in business services and the care home sectors. The Cleaning and Support Services Association indicated that ‘some 40% of members reported an initial compression of pay differentials… resulting in higher rates of pay for supervisory and management grades, … most of which was having to be absorbed as unobtainable from clients in price increases’. The Business Services Association reported that ‘inevitably, there has been some knock-on effect on differentials. These are of considerable importance in blue-collar work. The introduction of the NMW has tended to result in flattened pay structures’. The GMB said ‘there have also been reports of some erosion of differentials between care assistants and domestic staff as the minimum pay has been brought up’. And the Lancashire Care Association wrote that ‘in order to maintain pay differentials it has also been necessary in many cases to increase the hourly rate for more senior staff’.
A number of organisations associated the minimum wage with changes in the levels of pay at which staff could be recruited. One employer reported that ‘the National Minimum Wage is now considered to be "Poverty Level" and will continue to be considered so at whatever level it is set. To obtain competent staff employers must pay above this level — room must therefore be allowed for employers to build or enhance the National Minimum Wage in the light of their ability to do so’.
|
The general feeling amongst members is that their wage bills have increased by around 20% which is directly attributable to the introduction of the National Minimum Wage, and them having to maintain differential in pay levels for the various grades of staff. United Care Homes Association evidence |
We looked at the earnings distribution across the economy to see whether we could identify the impact on differentials. Figure 3.10 shows the increases in pay, excluding overtime earnings, across the earnings distribution between 1998 and 2000 using grossed NES data. These data show that the differential effect was contained within the bottom half of the earnings distribution, with the major effect at the eighth percentile or about £3.65 per hour (at 1998 earnings) and petering out by the 32nd percentile or around £5.35 per hour.
Figure 3.10
Percentile Increases in Hourly Earnings for Employees Aged 18 and over, 1998—2000
Source: Grossed NES data, April 1998 and 2000
Figure 3.11 disaggregates the data into changes between 1998/99 and 1999/2000. This suggests that the major impact between 1999 and 2000 occurred at around the eighth percentile (£3.90 per hour at 1999 levels of earnings) and peters out by around the 40th percentile (£6.20 per hour).
Figure 3.11
Percentile Increases in Hourly Earnings for Employees Aged 18 and over, 1998/99 and 1999/2000
Source: Grossed NES data, April 1998, 1999, 2000
We noted that the increases for 1999/2000 were lower than in 1998/99 and concluded that the differential effect of the National Minimum Wage appears to have become more muted since 1999 and that many organisations have returned to traditional across-the-board increases. The evidence for continuing restoration remains slight. IDS suggest that ‘any substantial changes to grading at the bottom of structures would have happened in year one of the NMW and subsequent pay reviews may have only required the straightforward application of a single percentage rise’ (IDS, 2000b).
IDS also analysed evidence on pay differentials in all firms which specified this as a pertinent issue (IDS, 2001a). It found these cases hard to come by, with isolated instances restricted to certain low-paying sectors such as textiles, retail, leisure, voluntary groups and hotels. Its evidence shows that most companies have not been forced to boost their lowest rates by higher percentages than those for supervisors. For example, at Roadchef Motorways, both probationers and supervisors received rises of 2 per cent from September 2000.
Our own survey of employers in low-paying sectors (that is, those firms we expected to be most affected by the introduction of the minimum wage) also provided us with information about the impact on differentials. Appendix 2 describes the survey methodology, sample and results more fully. As Table 3.4 shows, the data indicated that the median highest pay rate increased as a result of the National Minimum Wage was around £5 per hour.
| Table 3.4 | |||||
| Median Highest Hourly Rate Increased | |||||
| £ per hour | Hairdressing | Hospitality | Social care | Childcare | Total |
| Median | 5.00 | 5.28 | 5.00 | 4.95 | 5.00 |
|
Source: Low Pay Commission Survey, September—November 2000 |
|||||
Table 3.5 shows that for three-quarters of respondents the highest rate increased was less than £6 per hour, and for half of respondents it was less than £5 per hour. Our latest survey suggested that wages were increased higher up the distribution than we found in the survey we undertook for our second report. As noted in Appendix 2, the surveys are not directly comparable as they used a different sample in terms of size and composition.
| Table 3.5 | |||||
| Distribution of Highest Hourly Rate Increased | |||||
| £ per hour | Hairdressing (%) |
Hospitality (%) |
Social care (%) |
Childcare (%) |
Total (%) |
| £3.60 to £3.99 |
6
|
5
|
4
|
7
|
6
|
| £4.00 to £4.49 |
15
|
15
|
21
|
24
|
20
|
| £4.50 to £4.99 |
14
|
15
|
19
|
19
|
17
|
| £5.00 to £5.49 |
22
|
16
|
17
|
20
|
19
|
| £5.50 to £5.99 |
12
|
13
|
9
|
12
|
11
|
| £6.00 to £6.49 |
10
|
8
|
4
|
7
|
7
|
| £6.50+ |
21
|
28
|
25
|
12
|
21
|
|
Source: Low Pay Commission Survey,
September—November 2000 |
|||||
The definition of the minimum wage adopted in the National Minimum Wage Regulations allowed employers some latitude in restructuring their pay and working arrangements to take maximum advantage of the National Minimum Wage calculation. For example, employers could agree with their workers to consolidate certain pay elements that did not count towards the National Minimum Wage calculation, such as allowances and overtime premia, into basic pay so that these amounts counted towards the National Minimum Wage.
In our second report we stated that the ‘evidence at a sectoral level … shows that in general the implementation of the National Minimum Wage has gone very smoothly, and the definition of the wage has been easily accommodated by business’. Most of the evidence received for this report about pay structures related to differentials or to particular aspects of the National Minimum Wage Regulations, such as the accommodation deductor or tips (with which we will deal in the second volume of our report).
The latest overall picture on the composition of earnings for the low paid from the NES statistics is shown in Figure 3.12 below. It shows that low-paid workers continue to be less likely to receive additions to basic pay than higher-paid workers. Figure 3.13 shows the figures for those in receipt of overtime and incentive pay by sector. Overtime payments remain the major additional component for low-paid workers, followed by incentive and shift pay. As expected, incentive pay remains most important in the clothing and footwear sector, followed by retail.
Figure 3.12
Employees in Receipt of Additions to Basic Pay, 2000
Source: Grossed NES data, April 2000
Figure 3.13
Proportion of Employees in the Lowest Decile of Earnings in Receipt of Overtime and Incentive Payments in Low-paying Sectors, 2000
Source: Grossed NES data, April 2000
Figure 3.14 shows the proportion of employees in the lowest decile receiving additions to basic pay from 1998 to 2000 by sector. The data indicate that there was an overall decline in the percentage of employees receiving additional components for both the lowest decile and the rest since 1998. There was no clear evidence, however, of a minimum wage effect, as the picture is variable across the low-paying sectors. In some low-paying sectors there was an increase for the lowest paid, for example in clothing and footwear and in agriculture.
Figure 3.14
Proportion of Employees in the Lowest Decile of Earnings in Receipt of Additions to Basic Pay in Low-paying Sectors, 1998—2000
Source: Grossed NES data, April 1998, 1999, 2000
In summary, the earnings distribution has continued to narrow, as has the gender pay gap. The data suggest that the vast majority of workers received their entitlement to the National Minimum Wage. A recent development has been the creation of a ‘mezzanine’ earnings level at around £4.00 per hour, partly driven by the minimum wage and partly a result of tight labour market conditions. Our evidence suggests that the impact on differentials was limited, but some sectors were more affected than others. The major impact on differentials at the aggregate level appeared to occur with the introduction of the minimum wage; more recent changes were more muted. But the impact on firms in low-paying sectors was more marked. Evidence on pay structures shows that overtime continued to be the major additional earnings component for low-paid workers. And, in line with the trend for all workers, there has been a continuing decline in low-paid employees receiving additions to basic pay.
Employment and Non-employment
In this section we analyse the effect of the minimum wage on employment, unemployment and inactivity, focusing on the groups and sectors most affected by the minimum wage. In our second report we observed that there was no automatic link between employment and the minimum wage, and this section reviews the continuing evidence on this. We also explore the dynamic effects of the new legislation, the change in labour demand and supply on flows between employment and non-employment.
Employment
One of the conventional criticisms of minimum wages is that they cause job losses among those who stand to benefit from them. On the other hand the new economics of the minimum wage suggests that the impact on employment is broadly neutral. In this section we assess the extent to which these hypotheses are borne out by the evidence, both at the aggregate level and on groups most likely to be affected. We also present the results of econometric research we commissioned. Since this controls for factors other than the minimum wage which affect individuals’ employment prospects, if any negative employment effects are to be found we would expect them to show up here.
The period since the minimum wage was introduced has been one of strong employment growth and a tightening labour market. LFS employment (the number of people in work) was nearly 28 million in October to December 2000, which is a record level. Employment grew by 765,000 since the date of our first report, and by slightly over 430,000 since the minimum wage was introduced.
We saw in our analysis of beneficiaries of the minimum wage that the groups most affected were women, part-time workers, young people, those from ethnic minorities and workers with a work-limiting disability. As Figure 3.15 shows, the labour market performance of these groups has remained strong following the introduction of the minimum wage.
Figure 3.15
Change in Working Age Employment Rate of Different Groups, Spring 1999—Spring 2000
Source: LFS, Spring 1999, 2000
Between Spring 1999 and Spring 2000 the employment rate of working-age women grew fractionally less than that of men; 0.6 percentage points compared with 0.7 percentage points for men. The employment rate of young people grew at more than twice the aggregate. The increase in the employment rate among working-age ethnic minority groups was, on aggregate, much stronger than that of white workers, 1.3 percentage points compared with 0.7 for white workers. Increases were particularly marked for Black and Pakistani/Bangladeshi workers. The employment rate among working-age people with a work-limiting disability has increased by 0.5 per cent, slightly lower than the rise in the aggregate employment rate.
Over the year to Spring 2000, employment growth has been particularly marked for the 18—21 age group. As Figure 3.16 indicates, this has led them to increase their share of total employment, after controlling for relative population changes between the different age groups. This issue will be explored in more detail in volume two of our report. Part-time employment also continued to grow following the introduction of the minimum wage. By the end of 2000 there were around 7 million part-time workers, an increase of 2.3 per cent.
Figure 3.16
Change in Age-specific Employment Shares, Spring 1999—Spring 2000
Source: LFS data, Spring 1999, 2000
|
Low Pay Commission Research The Equality Commission for Northern Ireland conducted research for us on the impact of the minimum wage on the employment patterns of men and women in Northern Ireland. The research indicated that among firms in the survey direct employment losses seemed to have been relatively minor. A few firms attributed cuts in hours and job losses to the impact of the National Minimum Wage. But job losses claimed by firms to be attributable to the National Minimum Wage accounted for only about 1 per cent of total employment in those firms reporting a fairly significant or major effect from the minimum wage. The research findings emphasised that the minimum wage may have brought forward job losses that were going to occur anyway. Tony Dignan, 2001. The National Minimum Wage: Its Impact in Northern Ireland on the Employment Patterns of Men and Women |
While it is encouraging that the aggregate data do not show an adverse employment effect, they do not tell us what employment would have been without the minimum wage. In order to investigate this we need to establish a counterfactual. We commissioned research (Stewart, 2001) to investigate the impact of the introduction of the National Minimum Wage on the probability of employment among those whose wages were below the minimum wage prior to its introduction. We considered this to be the most robust way to identify possible negative employment effects of the minimum wage.
The approach adopted by Stewart was equivalent to other studies in the United States and France (see, for example, Abowd et al., 1999). The group directly affected by the minimum wage was compared with a similar group from just above the minimum, which is used as a counterfactual. The difference between the two groups was compared across periods that cover the introduction of the National Minimum Wage and earlier periods that do not. Control variables were also included in the analysis to account for other factors that could influence the probability of employment.
Results were presented for four demographic groups — male and female adults and youths, and for three different datasets — the LFS, NES and BHPS. Overall, the results suggest that the impact of the minimum wage on the probability of remaining in employment is zero or, if anything, marginally positive for adult men and young people. The conclusion for adult women was less clear-cut. There was weak evidence of a reduction in employment probability post-minimum wage, but even for this group the result was not significantly different from zero.
We considered this to be a powerful addition to the growing body of evidence on the employment effects of minimum wages (see for example OECD, 1998 for a recent summary). It suggested that the orthodox view that minimum wages cut employment may not be appropriate when analysing low pay in the UK, and that other explanations, such as imperfections in the labour market, may be more relevant.
These aggregate trends mask considerable variations between sectors and between regions. Figure 3.17 shows that employment in security has increased relatively quickly, while employment in the hospitality and retail sectors has increased in line with the national average. Employment in textiles and cleaning has declined in accordance with longer-term trends and structural changes in these sectors.
Figure 3.17
Employment Change in Low-paying Sectors, 1997—2000
Source: Employee jobs series, ONS, 1997—2000
|
Overall we have not found negative employment effects arising as a consequence of the introduction of the NMW. Here and there we have found companies with a little local difficulty, but with a bit of ingenuity most problems have been solved. A look at the official figures on employment growth in 1999/2000 shows growth in a range of services, with the largest growth in distribution, hotels and catering. IDS, 2000b |
Sectors experiencing longer-term structural change are more at risk of external shocks such as the minimum wage. All the evidence we looked at — our own survey, external research and written submissions — showed that the minimum wage particularly affected firms in low-paying sectors and where competitive pressures were strongest. The impact on employment depended on a number of factors, principally the ability of firms to reduce profits and increase prices. A survey we commissioned, of firms affected by the minimum wage in Northern Ireland (Dignan, 2001), found that textiles and clothing firms accounted for 73 per cent of the reported direct job losses, and that employers in this sector were generally unable to raise prices in adjusting to the minimum wage.
Of the firms in our survey which reported that the minimum wage had affected their businesses, around 13 per cent said that they had made significant reductions to their staffing levels as a result and 11 per cent reported that they had reduced working hours significantly. There may be a displacement effect that is reducing employment in individual sectors and firms, which is offset by rises in others. This would reconcile these findings with the aggregate level result of continued employment growth. The employment effects varied considerably by sector as illustrated in Figure 3.18. In general, the hairdressing, textile and hospitality sectors were more likely to report reductions in hours and staff, while social care and childcare were less likely. By contrast, these last two sectors were more likely to report reductions in profits and increases in prices.
Figure 3.18
Reductions in Staffing Levels as a Result of the National Minimum Wage
Source: Low Pay Commission Survey, September—November 2000
Note: Base, all firms affected by the National Minimum Wage in any way.
|
The GMB reported that the clothing and textiles industry had suffered a huge number of redundancies over the last year. The main influence on these was overseas competition, and the decision of the main retailer, Marks & Spencer, to reduce the volume of clothing supplied from the United Kingdom. GMB evidence |
The British Retail Consortium’s own survey showed that while no large businesses had reduced staff as a result of the minimum wage, 11 per cent of small retailers had done so. The Association of Convenience Stores’ survey indicated that 44 per cent of respondents said that the minimum wage had led to a reduction in the number of staff employed.
Reduced working hours were reported by the Association of Convenience Stores, the Biscuit, Cake, Chocolate and Confectionery Alliance, the Brewers and Licensed Retailers Association and the Hairdressing Employers Association. The Amalgamated Engineering and Electrical Union (AEEU) evidence indicated that some employers were not recording the full hours worked in order to reduce their National Minimum Wage liability. Figure 3.19 shows the reduction in working hours by sector as reported in our own survey.
Figure 3.19
Reductions in Working Hours as a Result of the National Minimum Wage
Source: Low Pay Commission Survey, September—November 2000
Note: Base, all firms affected by the National Minimum Wage in any way.
The reduction in hours identified in certain sectors is borne out to some extent by NES data on hours worked. The latest figures for paid hours of work show that while there has been little change overall, with total hours, including overtime, for the low paid averaging around 27 per week compared with 36 for the rest, the story was different for some sectors most affected by the National Minimum Wage. The sectoral data indicate that both basic and overtime hours fell for lowest-decile workers in retail, hospitality, agriculture and cleaning.
A number of organisations pointed to a regional impact. The Brewers and Licensed Retailers Association said that introducing the minimum wage at a sensible level had served to lessen the impact on some areas of the UK. Nevertheless, members reported that there had still been some regional impact, particularly in the North and the South West of England. The British Hospitality Association mentioned particularly the impact in Scotland, the South West and Wales.
If there were to be an employment effect, we would expect the aggregate data to reflect regional differences, with a minimum wage effect more evident in lower-paid regions. As Figure 3.20 below shows, this did not appear to be the case.
Figure 3.20
Distribution of Low Pay in 1998 and Change in Employment and Unemployment Rates Post-National Minimum Wage by Region
Source: LPC calculations based on grossed NES and LFS data for earnings
at Spring 1998. LFS data for employment and unemployment at Spring
1999 and Spring 2000
Note: Government Office Regions shown for England. Data for Northern Ireland
are not seasonally adjusted.
|
Low Pay Commission Research We commissioned research to examine the effect of the minimum wage on small and medium-sized enterprises in Yorkshire and Humberside. The study found that the three most important means of offsetting labour costs were increasing prices, reducing hours and improving product or service quality. When asked to list the three most important methods of offsetting labour costs, the factor most frequently mentioned by respondents was reducing working hours. Fifty-eight per cent of respondents viewed this as being one of the three most important methods. Heyes and Gray, 2001. The Implications of the National Minimum Wage for Training and Employment in Small Firms |
London was the region with the lowest proportion of employees earning below the minimum wage in 1998, but it had a small fall in employment in the post-minimum wage period. The lowest-paying region was the North East, but this had the largest increase in employment and the largest fall in unemployment post-minimum wage. This suggested that factors other than the National Minimum Wage are more important in determining regional employment and unemployment rates.
The trend in unemployment mirrored, to a large extent, the trend in employment post-minimum wage. International Labour Organisation (ILO) unemployment stood at 5.3 per cent in October to December 2000, the lowest level since the ILO measure of unemployment was introduced in 1984. ILO unemployment fell by just under 235,000 since the minimum wage was introduced.
As with employment, unemployment falls have been broadly based: unemployment is now lower than the previous trough in 1990 across most age groups and among both males and females. Figure 3.21 shows that unemployment rates fell in all age groups since the minimum wage was introduced, apart from 16—17 year olds and 22—24 year olds. For the former group the rise in unemployment was concentrated among those in full-time education. As participation in education has increased substantially among this group, so has the proportion that is seeking work. For the latter group, the Spring 2000 to Autumn 2000 rise in unemployment appears to be due to seasonal factors related to leaving full-time education. The proportion of this age group in full-time education declined by 2.7 per cent over this period, with many of this group still seeking work at Autumn 2000.
Figure 3.21
ILO Unemployment Rates by Age, 1993—2000
Source: LFS data, Spring and Autumn 1993—2000
In the year after the minimum wage was introduced the ILO unemployment rate fell faster for men than for women, 0.7 percentage points compared with 0.3 for women. Some of this can be explained by the larger rise in participation among women than among men. Unemployment also fell across different durations. In a period of rising employment, long-term unemployment (those unemployed for a year or more) might be expected to fall as fewer short-term unemployed reach this duration. This has been the case since the minimum wage was introduced, with long-term unemployment falling by 21 per cent compared with 10 per cent among those unemployed for up to six months.
There was also a continued improvement in the unemployment situation of ethnic groups and people with a work-limiting disability post-minimum wage. Recent unemployment trends by ethnic group are shown in Figure 3.22. Between Spring 1999 and Autumn 2000, the ILO unemployment rate for white workers fell by 0.6 per cent. The fall among the other ethnic groups shown in Figure 3.22 has been faster — between 1 and 1.3 per cent. The fall in unemployment has also been marked for people with a work-limiting disability — 1.6 per cent over this period.
Figure 3.22
ILO Unemployment Rates by Ethnic Group, 1993—2000
Source: LFS data, Spring and Autumn 1993—2000
By increasing the earnings of the low paid, and giving clarity to the level of earnings people can expect in work, the minimum wage might be expected to improve work incentives for those moving into low pay and increase participation (employment plus unemployment) in the labour market. Figure 3.23 shows changes in labour market participation in the year following the introduction of the minimum wage. Participation has risen for all ages apart from 22—24 year olds.
Figure 3.23
Change in Participation Rates by Age, Spring 1999 —Spring 2000
Source: LFS data, Spring 1999, 2000
Disaggregating changes in inactivity by gender produced some mixed results. Men and women aged 45—49 increased their inactivity post-minimum wage as did women aged 20—24. The reasons for inactivity are unlikely to be minimum wage-related. Among the former group, a greater proportion stated they were long-term sick. For the latter group, increased inactivity was due to a move from unemployment into inactivity as proportionately more women chose to look after the family/home, which more than offset higher employment among this group.
We also analysed movements from inactivity to other labour market states using the linked LFS data. Figure 3.24 gives the percentage of working-age inactive people leaving inactivity from one quarter to the next. It shows that in the post-minimum wage period the proportion leaving inactivity was broadly flat.
Figure 3.24
Percentage Leaving Inactivity from One Quarter to the Next by Age (4 Quarter Moving Average), 1993—2000
Source: LFS Longitudinal data, 1993—2000
Note: Due to measurement error, proportions of people changing status
are likely to be biased upwards, while proportions not changing status
are likely to be biased downwards.
Data on aggregate flows mask the degree of labour market attachment of the inactive group. Over the year since the minimum wage was introduced, the number of inactive people who said they would like a job remained broadly stable, although this was against a background of a fall in the working-age inactive population of around 75,000. This means that the proportion of the inactive who said they would like a job increased. Among the group showing the largest propensity to move closer to the labour market are discouraged workers (people who wanted a job but were not seeking because they believed no jobs were available); those looking after the family or home; and students.
|
Each year the Greater Manchester Low Pay Unit conducts a detailed survey of Jobcentre vacancies in each of the ten major Jobcentres in the county. The April 2000 Jobcentre survey data show a rise of 11.7 per cent on the number of jobs recorded in April 1999. This suggests that the introduction of the minimum wage has not had a negative impact on the number of jobs available in this particular labour market. Greater Manchester Low Pay Unit evidence |
The incentive effects of the minimum wage were analysed in more detail in research we commissioned (Stewart and Swaffield, 2001) using the BHPS data. This considered job prospects for the unemployed post-minimum wage. The analysis showed that nearly 20 per cent of unemployed respondents said that the range of jobs that they would be prepared to take has widened because of the National Minimum Wage. This effect was slightly greater for women than for men. This, together with the fact that the unemployed generally regarded their job prospects to be unchanged since the minimum wage was introduced, provided support for the argument that the minimum wage has had a positive impact on labour supply and widened the range of jobs people moving into low pay are prepared to take.
The minimum wage has to be considered together with other labour market interventions aimed at making work pay. The second volume of our report will consider the interaction of the National Minimum Wage with the tax and benefits system in more detail.
|
Low Pay Commission Research A survey of students by Rosemary Lucas and Michelle Langlois, which we commissioned, suggested that the minimum wage itself had little impact on the decision by students to work. Most would seek part-time work whether or not the minimum wage applied. Lucas and Langlois, 2000. The National Minimum Wage: What Can Young Workers Tell Us? |
We would expect higher wages to improve recruitment and retention within firms. Labour turnover is strongly correlated with employment growth, with periods of high employment growth characterised by periods of high turnover. In the post-minimum wage period of high employment growth, we might therefore expect to see increased labour market turnover. Figure 3.25 shows the percentage of employees remaining in employment from one quarter to the next. Contrary to expectation it suggests that duration of employment has increased over the last few years. Between Winter/Spring 1998 and Winter/Spring 2000, the proportion of those aged 22 and over remaining in employment from one quarter to the next rose from 73 per cent to 74.2 per cent.
Figure 3.25
Percentage of Employees Remaining in Employment from One Quarter to the Next by Age (4 Quarter Moving Average), 1993—2000
Source: LFS Longitudinal data, 1993—2000
Note: Due to measurement error, proportions of people changing status
are likely to be biased upwards, while proportions not changing status
are likely to be biased downwards.
Labour turnover is more prevalent in low-paying sectors and firms. Evidence from research we commissioned into small and medium-sized enterprises (Bullock et al., 2001) showed that of firms with at least half their workforce paid at the minimum, 16 per cent had turnover of less than 10 per cent compared with 44 per cent of firms paying above the minimum. Addressing labour turnover is therefore more important in the lowest-paying firms.
Our survey asked whether firms affected by the minimum wage had experienced any benefits to their business. Almost 30 per cent of firms reported that they had experienced some benefit, but few stated any significant benefit. As can be seen from Figure 3.26, the most common benefit reported was higher staff motivation, indicated by over 20 per cent. Over 10 per cent reported lower staff turnover and 8 per cent the faster filling of vacancies. Textiles firms were least likely to report any benefit to business (only 11 per cent), with none indicating any significant benefit. The results from the textiles sector were not surprising given the global competitive pressures the sector is currently facing.
Figure 3.26
Benefits to Business as a Result of the National Minimum Wage
Source: Low Pay Commission Survey, September—November 2000
Note: Base, all firms affected by the National Minimum Wage in any way.
Evidence we received suggested that the effects of the interaction of the minimum wage with the tight labour market were often difficult to disentangle. The National Pharmaceutical Association reported that ‘over the sector as a whole, the introduction of the minimum wage does not appear to have impacted negatively on recruitment or retention policies of our members. However, in some areas, particularly the South East and other areas where wage levels are above average, members are experiencing difficulties in recruitment’.
Research conducted for us (Heyes and Gray, 2001) found that small employers’ ability to recruit had tended to worsen (13 per cent) rather than improve (2 per cent). And retention had worsened in 9 per cent of firms, and improved in 4 per cent. In both cases increased competition in the labour market was cited as the reason.
A common theme among some firms has been a tendency to increase wages above the minimum to aid retention and recruitment in a tight labour market. Research by the BRC showed that a quarter of retailers had increased pay rates as a result of the minimum wage, even though they were already paying more than the relevant rate. Small businesses were more likely to increase rates than larger. The BRC said that this was because larger firms could offer better terms and conditions and career progression to retain staff. Interestingly, Heyes and Gray (2001) found evidence that some small firms recognised non-wage benefits, such as training, as a means to retain staff.
|
As labour markets have tightened over the past 18 months, firms with wages in the £3.60 to £4.00 range have experienced some difficulties in staff recruitment and retention. Some firms are finding that they are having to re-position themselves in relation to others. Some lower paying firms have been rather surprised to find which other firms they are now in competition with, outside of their previous expectations. IDS, 2000a |
IDS found this trend was mirrored in other sectors. The minimum wage led to greater competition between high-street firms which has in turn led to basic rates being increased to between £4.00 and £4.50 or higher. One textiles company told IDS that ‘we set wages at this level (£4.70) to stay ahead of other firms in the area who were mostly paying the minimum wage’ (IDS, 2000b).
In summary, the labour market has continued to tighten post-minimum wage. Employment continued to grow strongly, unemployment fell further and participation in the labour market increased. We were particularly concerned about employment among those groups most affected by the minimum wage: women, young people, ethnic groups, part-time workers and disabled workers. Aggregate data show that, in the main, these groups have shared in the overall growth in employment. But in a strong labour market negative employment effects could be difficult to detect, and employment may have been even higher in the absence of a minimum wage. Results from research which control for this and other factors were encouraging, and show that the effect of the minimum wage was broadly neutral or, if anything, mildly positive. Firms can make a range of adjustments to control labour costs other than shedding labour, and there was some evidence of firms reducing hours post-minimum wage; but these tended to be confined to a minority of firms in particular sectors.
Despite the minimum wage, many low-paying sectors faced retention and recruitment difficulties. The strong labour market meant that some firms faced competition from firms and sectors outside of their previous experience. The minimum wage had a mild, positive impact on work incentives. The minimum wage, together with other active labour market policies, could be helping to increase labour supply in a tight labour market.
