The impact of the National Minimum Wage on the by idv45773



         The impact of the National Minimum Wage on the
            pay distribution, employment and training

                                 David Metcalf
                           London School of Economics
                                   May 2003
                              Revised August 2003


Britain’s first ever national minimum wage (NMW) came into force in April 1999
covering some 5% of workers. Between 1999 and 2002 the NMW rose in line with
the growth in average earnings. But for 2003-06 the NMW will be ratcheted-up
relative to median pay. The evidence shows that between 1998-2002: first, the
bottom decile of the pay distribution experienced above average pay rises with no
spillover effects further up the distribution. Second, there was no overall employment
effect, but a small negative impact in the care home sector. Third, the NMW boosts
the probability and intensity of training.

Attention Printer
Pagehead Title: National Minimum Wage

JEL Classification: J5

Key words: National Minimum Wage

Disclaimer: I am a member of the Low Pay Commission (LPC) but am writing here
in a personal capacity: the paper should not be taken as representing views of the LPC
or its secretariat.

         Britain’s first ever national minimum wage (NMW) came into force in April

1999. The adult hourly rate was £3.60 with a slightly lower rate for those aged 18-21

(see Table 1). The Low Pay Commission (LPC), which recommends the rate to the

government, anticipated that some 1.8m jobs would be covered by the introduction of

the NMW. In the event, because the statistics on low paid employment supplied to

the LPC by the Office of National Statistics were seriously deficient, coverage turned

out to be around a third lower at 1.2 million.

         Between 1999 and 2002 the NMW rose roughly in line with the average

earnings index (AEI). But in 2003 and 2004 the hike in the NMW is double the

predicted rise in the AEI. The 2004 adult rate is £4.85 an hour. Further the LPC has

recommended, subject to economic conditions, the minimum wage to rise faster than

the AEI in 2005 and 2006. A major reason that the LPC is confident in

recommending a gradual ratcheting-up in the real value of the NMW is because the

evidence (LPC 2003, chapter 2) suggests that the NMW has resulted in a boost in pay

for those at the bottom of the pay distribution without any corresponding adverse

impact on aggregate employment.

         The four outstanding papers, presented at the RES April 2003 conference and

discussed below, confirm this earlier evidence. The papers address three main NMW

issues: its impact on the distribution of pay; associated employment changes, in

aggregate and in low paid sector; and links with the incidence and extent of training –

a key component of the rather neglected area of “competitiveness” which also covers

productivity and prices.

         LPC evidence (2003, figure 2.14) shows that between 1998 and 2002 the

bottom decile of the pay distribution experienced pay rises well above the median

increase (with the effect tapering from a large boost for the bottom percentile down to


a similar to median increase at the 11th percentile), but with no spillover effects

further up the distribution. This impact of the NMW on the overall pay distribution is

confirmed in the care homes sector (Dickens and Manning 2002, Machin and Wilson

2003) which is the lowest paying sector of the UK labour market. All UK care homes

were surveyed by post both before and after the introduction of the NMW in 1999 and

after the 2001 uprating. Prior to the introduction of the NMW some 40% of workers

in this sector were paid below £3.60. Employers did not anticipate the NMW such

that pay gradually rose in the months prior to its introduction. Rather there was what

Dickens and Manning call a “big bang” in April 1999, resulting in a significant spike

in the care home pay distribution at £3.60. This lack of any anticipation effect is an

important finding because any such effect could be a reason why coverage of the

initial NMW was lower than predicted. In the event this is not the case, which

confirms ONS data deficiencies were the culprit. The evidence also points to almost

complete compliance with the NMW, but this could be because non-compliers were

less likely to return the questionnaire.

         Information from care homes also confirms the lack of spillover effects further

up the pay distribution. This is a double-edged result. It implies that the NMW had a

more modest impact on tempering wage inequality than it otherwise might have done.

But, on the other hand, any inflationary consequences are also limited, with a more

benign effect on interest rates than would have occurred if there had been wage

spillovers much further up the distribution.

         Employment changes consequent on the introduction and uprating of the

NMW have been studied intensively: in aggregate; by industry and age shares; by

individual characteristics; by region; and in the low paying care home sector. The


evidence suggests no overall employment effect of the NMW but a small negative

impact in the care home sector.

         Stewart (2002) builds on his earlier work (see LPC 2003, table 2.5) to analyse

the employment effects of the 2000 and 2001 upratings. The Labour Force Survey

and New Earnings Survey Panel are used to undertake both difference-in-difference

and wage gap (which examines by how much the individual’s wage was raised)

estimates by age, gender and sector. An important innovation in this paper is the

analysis of any differential impact of the slowdown in aggregate employment growth

during this period. Overall, the results indicate no adverse employment effects

associated with the upratings. There is a hint of a negative effect for adult women

“but it is sensitive to the choice of wage variable and estimation method” and, further,

for female-intensive sectors of employment “no evidence of a negative effect is found

for the retail sector, for cleaners, or for child care workers and care assistants”.

         If the NMW has employment effects a particular sector where we would

expect to find such evidence is care homes. This is because so many workers were

paid below the NMW prior to its introduction, subsequently there is a big spike in the

care home wage distribution at the NMW, and because cost increases cannot easily be

passed on: a substantial fraction of consumers are paid for by the state at a rate

negotiated well in advance. Machin and Wilson (2003) find no evidence that homes

most heavily affected by the introduction of the NMW were more likely to close

down in the two and a half year period after the introduction of the wage floor. But

they do find “moderate” employment effects, with an elasticity of employment with

respect to the wage (increase above the average) of around -0.3. But even this modest

negative elasticity may reflect workers switching from low-paying to high-paying

care homes.


         The effect of a NMW on the incidence and extent of training had not been

studied until Arulampalam et al.’s thorough study (2002). If the labour market is

competitive it is predicted that the minimum wage makes it less profitable to employ

unskilled workers. But in a non-competitive labour market the firm is getting a rent

and so would like to retain the worker – and the firm now has an incentive to improve

the productivity of the employee via training in order to restore the surplus. Using a

longitudinal difference-in-difference method and data from BHPS 1998-2000 the

authors demonstrate that the probability of training incidence and the intensity of

training increased by 8-11 percentage points for affected workers. Like the lack of

overall employment effects, this finding suggests that the competitive model may not

be a complete description of the low paid labour market.

         Both the introduction and uprating of the NMW permit quasi-experimental

research into its effects. These four salient papers adopt such a method to inform us

about major issues – the effect on the wage distribution, employment and training.

Such academic research has been, and continues to be, the vital input in helping the

LPC to decide what rate(s) to recommend in future. Such constructive interactions

between the academic community and policy-makers are a key ingredient in the

successful introduction of the NMW in the UK.

London School of Economics


                                        Table 1
                          UK Hourly National Minimum Wage (£)

                                   Adult rate               Development rate*
   April 1999                      3.60                     3.00
   June 2000                       3.60                     3.20
   October 2000                    3.70                     3.20
   October 2001                    4.10                     3.50
   October 2002                    4.20                     3.60
   October 2003                    4.50                     3.80
   October 2004                    4.85                     4.10

   Note: * denotes rate for employees aged 18 to 21 and those aged 22 and over receiving
   accredited training and in the first six months of employment.



Arulampalam, W., Booth, A. and Bryan, M. (2002) ‘Work-related training and the
      new national minimum wage in Britain’, mimeo, Essex University, August.

Dickens, R. and Manning, A. (2002) ‘Spikes and spillovers: the impact of the national
      minimum wage on the wage distribution in a low wage sector’, mimeo, CEP,
      LSE, December.

Low Pay Commission (2003) The National Minimum Wage. Fourth Report of the
      Low Pay Commission, Cm 5768, March, London: Stationery Office.

Machin, S. and Wilson, J. (2003) ‘Minimum wages in a low wage labour market: care
      homes in the UK’, mimeo, CEP, LSE, April.

Stewart, M. (2002) ‘Modelling the employment effect of the minimum wage: revised
       report to LPC’, mimeo, Warwick University, November.


To top