Table of Contents
Type: Research Paper | Subject: Economics | Level: Masters | Word Count: ~3,200 words | Referencing: Harvard
This model research paper was produced by an Essays UK specialist as reference material for learning purposes only. For support in this field, see our economics assignment specialists.
For your MSc Labour Economics module, write a research paper (3,000–3,500 words) critically reviewing the evidence on the employment effects of the UK National Living Wage, and propose an appropriate empirical strategy for estimating these effects using a difference-in-differences framework.
This paper examines the employment effects of the UK National Living Wage (NLW), introduced in April 2016 as a higher statutory wage floor for workers aged 25 and over. Using a stylised difference-in-differences framework applied to illustrative regional and sectoral data, the paper estimates the impact of the NLW’s introduction on employment, hours and wage growth in low-paying sectors such as retail and hospitality. Consistent with recent UK evidence, the analysis finds no statistically significant negative effect of the NLW on aggregate employment, alongside a significant positive effect on wage growth for workers at or near the wage floor. Some evidence of a small, statistically significant reduction in average hours is identified, suggesting a degree of adjustment on the hours margin rather than through job losses. These findings are discussed in relation to monopsony models of the low-wage labour market, which predict smaller, and potentially positive, employment effects of minimum wage increases than the standard competitive model. The paper concludes that, within the range of increases implemented to date, the NLW has raised pay for low-paid workers without generating the substantial job losses predicted by simple competitive theory, though continued monitoring is warranted as the wage floor rises relative to median pay.
The UK National Living Wage (NLW) was introduced in April 2016 as a new, higher minimum wage rate for workers aged 25 and over, set at a level intended to reach 60 per cent of median hourly earnings, a target subsequently extended and, in later years, broadened to cover younger workers too (Low Pay Commission, 2022). The policy represented a significant departure from the previous National Minimum Wage (NMW) regime, both in the size of the increase it implied for many low-paid workers and in the more ambitious, median-linked target that governed its future path. Because the NLW was set with reference to median earnings rather than being fixed solely by reference to employment effects, it raised long-standing questions in labour economics about how large minimum wage increases can be before they begin to reduce employment among low-paid workers.
The standard competitive model of the labour market predicts that a wage floor set above the market-clearing wage will reduce employment, as firms respond to higher labour costs by hiring fewer workers or reducing hours. This prediction has historically informed a cautious approach to minimum wage policy. However, an influential body of empirical research, beginning with Card and Krueger’s (1994) study of fast-food employment in New Jersey, has repeatedly failed to find the employment losses predicted by the competitive model, and in some cases has found small positive employment effects following minimum wage increases. This body of evidence has contributed to the development of monopsony-based models of the low-wage labour market, in which employers hold some wage-setting power over workers, for example because of search frictions or limited outside options. Under monopsony, a minimum wage increase within a certain range can increase both wages and employment simultaneously, rather than trading one off against the other (Manning, 2021).
The introduction and subsequent above-inflation increases in the UK NLW provide a valuable natural experiment for testing these competing predictions in a European, sector-diverse labour market. The UK evidence base to date, including work commissioned by and reported to the Low Pay Commission, has generally found limited evidence of negative employment effects from NLW increases (Dickens, Riley and Wilkinson, 2015; Riley and Bondibene, 2017), broadly consistent with the international minimum wage literature reviewed by Cengiz et al. (2019) and Manning (2021). However, questions remain about whether these findings will continue to hold as the NLW rises further relative to median pay, and about how effects are distributed across sectors, regions, and margins of adjustment such as hours, non-wage benefits and prices.
This paper reviews the theoretical and empirical literature on minimum wage employment effects, with a particular focus on evidence relevant to the UK NLW, and sets out an empirical strategy, based on a difference-in-differences design exploiting variation in ‘bite’ (the proportion of workers affected) across sectors, for estimating the policy’s employment effects. Using stylised illustrative data constructed to reflect patterns reported in the UK evidence base, the paper then estimates and discusses the employment, hours and wage effects of the NLW’s introduction.
The theoretical starting point for most minimum wage research is the standard competitive labour market model, in which a wage floor set above the equilibrium wage necessarily reduces employment by moving the market up a downward-sloping labour demand curve (Neumark and Wascher, 2008). Early UK evidence broadly supported this framework: Machin and Manning’s (1994) study of the Wages Councils, sector-specific minimum wage bodies that operated in the UK before national minimum wage legislation, found that abolition of council-set minimum wages was associated with increased wage dispersion, consistent with binding wage floors compressing the wage distribution, though the employment effects identified were modest.
The empirical picture shifted substantially following Card and Krueger’s (1994) comparison of fast-food employment in New Jersey, where the state minimum wage rose, and neighbouring Pennsylvania, where it did not. Contrary to competitive theory, they found no evidence that the New Jersey minimum wage increase reduced employment relative to Pennsylvania, and some evidence of a small employment increase. This study prompted a wave of research using similar quasi-experimental designs. Dube, Lester and Reich (2010) extended the approach across US state borders more broadly, comparing counties on either side of state lines with different minimum wages, and again found no evidence of negative employment effects, while addressing earlier criticisms that Card and Krueger’s original comparison states were not representative.
More recently, Cengiz et al. (2019) used a ‘bunching’ estimator applied to the full distribution of US wages to directly estimate job losses at the bottom of the wage distribution following minimum wage increases, and found that job losses below the new minimum were closely offset by job gains just above it, implying a near-zero net employment effect even for relatively large increases, up to around 59 per cent of the local median wage. This finding is particularly relevant to the UK NLW, which was designed with reference to a similar proportion of median earnings.
UK-specific evidence has generally reached similar conclusions. Butcher, Dickens and Manning (2012) examined the effect of the National Minimum Wage on wage inequality and found substantial compression of the lower wage distribution with limited evidence of disemployment effects. In preparatory analysis for the Low Pay Commission ahead of the NLW’s introduction, Dickens, Riley and Wilkinson (2015) reviewed the international and UK evidence and concluded that, at the rates then proposed, the NLW was unlikely to generate large aggregate job losses, though they cautioned that effects could vary by sector and region depending on local ‘bite’. Riley and Bondibene (2017) examined firm-level productivity responses to UK minimum wage increases and found evidence that some firms respond to higher wage floors partly through productivity improvements rather than through headcount reductions, a channel not captured by simple competitive models. The Low Pay Commission’s own annual monitoring reports have consistently found limited evidence of negative employment effects from NLW upratings, while noting some evidence of adjustment through reduced hours, slower wage growth further up the distribution, and, in some periods, reduced non-wage benefits (Low Pay Commission, 2022).
The theoretical explanation for these findings has increasingly centred on monopsony models of the low-wage labour market. In a monopsonistic market, individual employers face an upward-sloping labour supply curve, meaning that to hire additional workers they must raise wages not only for new hires but, in many formulations, for existing staff too. This gives employers an incentive to pay below the competitive wage and hire fewer workers than would be socially efficient. A minimum wage introduced into such a market can increase both wages and employment up to the point where the wage floor reaches the competitive wage; only once it is pushed further does the standard disemployment prediction re-emerge (Manning, 2021). Search frictions, limited job mobility, and imperfect information about outside wage offers are commonly cited sources of employer wage-setting power consistent with this framework, and are plausible features of many UK local labour markets, particularly outside major cities.
Taken together, the literature suggests that the standard competitive prediction of substantial job losses from minimum wage increases has not been well supported empirically, either internationally or in the UK, at least for the range of increases implemented to date, and that monopsony-based models provide a more consistent account of the observed pattern of wage gains without matching job losses. However, most of this evidence relates to increases up to the levels seen historically; the continued rise of the NLW towards, and potentially beyond, prior benchmarks of median-wage ‘bite’ makes ongoing monitoring and further empirical work important.
Empirical strategy. This paper adopts a difference-in-differences (DiD) design, which is the dominant approach in the empirical minimum wage literature reviewed above (Card and Krueger, 1994; Dube, Lester and Reich, 2010). The core intuition of the DiD approach is to compare employment changes in areas or sectors more strongly affected by a minimum wage increase (‘high-bite’ areas, where a larger share of workers earn close to the new wage floor) with employment changes in areas less affected (‘low-bite’ areas), before and after the policy change. Under the assumption that high-bite and low-bite areas would have followed parallel employment trends in the absence of the NLW (the ‘parallel trends’ assumption), any divergence in employment growth between the two groups following the policy’s introduction can be attributed to the NLW itself.
Measuring bite. Following standard practice in this literature, sectoral bite is measured as the proportion of employees in a given sector earning below the incoming NLW rate in the period immediately prior to its introduction. Sectors such as retail, hospitality and social care typically show high bite, since a large share of their workforce is paid at or close to the minimum wage; sectors such as finance and professional services show very low bite, providing a natural comparison group.
Model specification. The baseline specification takes the form:
Est = α + β(Bites × Postt) + γs + δt + εst
where Est is an employment outcome (for example, the log of employment or average hours) in sector s at time t; Bites is the sector’s pre-policy bite measure; Postt is an indicator equal to one in periods after the NLW’s introduction; γs and δt are sector and time fixed effects, controlling for time-invariant sector characteristics and economy-wide shocks common to all sectors; and β is the coefficient of interest, capturing the differential employment effect in high-bite relative to low-bite sectors following the policy change. A negative and significant β would indicate that the NLW reduced employment relatively more in more affected sectors, consistent with the competitive model; a null or positive β would be more consistent with monopsony-based predictions.
Data. For the purposes of this illustrative analysis, quarterly sector-level employment, hours and average pay data were constructed for six low-to-moderate bite sectors over eight quarters spanning the NLW’s introduction (four quarters before, four after), calibrated to be broadly consistent with published Low Pay Commission and Office for National Statistics series rather than drawn directly from disclosed microdata.
Estimation and inference. The model was estimated by ordinary least squares with standard errors clustered at the sector level, to allow for correlated shocks within sectors over time. As a robustness check, an event-study specification, interacting bite with quarter-specific indicators, was used to examine whether high- and low-bite sectors showed similar employment trends prior to the NLW’s introduction, providing a test of the parallel trends assumption underlying the main specification.
Table 1 reports illustrative pre- and post-NLW employment, hours and pay outcomes for high-bite and low-bite sectors. High-bite sectors saw average hourly pay for the lowest-paid decile rise by 8.4 per cent in the year following the NLW’s introduction, compared with 2.1 per cent in low-bite sectors, consistent with the NLW compressing the lower half of the wage distribution as intended. Employment in high-bite sectors grew by 1.2 per cent over the same period, only marginally below the 1.6 per cent growth recorded in low-bite sectors, a gap that was not statistically significant. Average weekly hours in high-bite sectors fell slightly, from 29.8 to 29.1 hours, a larger decline than the 0.2-hour fall recorded in low-bite sectors.
| Outcome | High-Bite Pre | High-Bite Post | Low-Bite Pre | Low-Bite Post |
|---|---|---|---|---|
| Hourly pay, lowest decile (£) | 6.80 | 7.37 | 9.10 | 9.29 |
| Employment index (base = 100) | 100.0 | 101.2 | 100.0 | 101.6 |
| Average weekly hours | 29.8 | 29.1 | 31.4 | 31.2 |
Table 1. Illustrative sector outcomes, four quarters before and after NLW introduction.
The difference-in-differences estimate of the effect of bite on employment growth was β = -0.031 (SE = 0.028, p = .29), indicating a small and statistically non-significant negative association between sectoral bite and employment growth following the NLW’s introduction. By contrast, the equivalent estimate for hourly pay growth was β = 0.61 (SE = 0.11, p < .001), confirming a strong and significant wage effect concentrated in high-bite sectors, as expected given that the NLW directly raises the wage floor. For average weekly hours, the DiD estimate was β = -0.38 (SE = 0.17, p = .038), suggesting a small but statistically significant reduction in hours in more affected sectors, indicating that some adjustment to the NLW occurred on the hours margin even though overall headcount was not significantly affected.
Figure 1. Indexed employment (base = 100 at NLW introduction) in high-bite versus low-bite sectors, four quarters before and after introduction.
Figure 1 illustrates these patterns, plotting indexed employment for high-bite and low-bite sectors across the eight quarters surrounding the NLW’s introduction. The two series track one another closely in the four quarters before introduction, providing some support for the parallel trends assumption, and continue to move broadly together afterwards, consistent with the small and non-significant employment coefficient reported above. The event-study specification, examining bite-by-quarter interactions, found no statistically significant pre-trend differences between high- and low-bite sectors in the four quarters prior to the NLW’s introduction (all p > .10), supporting the validity of the parallel trends assumption underpinning the main difference-in-differences estimates. Post-introduction coefficients showed no evidence of a growing employment gap between high- and low-bite sectors over the two years following the policy change.
The illustrative analysis presented here finds no statistically significant negative employment effect of the National Living Wage on more affected, high-bite sectors relative to less affected sectors, alongside a strong and statistically significant positive effect on hourly pay for low-paid workers in those sectors. This pattern is consistent with the broader empirical literature reviewed above, from Card and Krueger’s (1994) original New Jersey study through to Cengiz et al.’s (2019) bunching estimates and UK-focused work commissioned for the Low Pay Commission (Dickens, Riley and Wilkinson, 2015; Riley and Bondibene, 2017). It is not consistent with the strong disemployment prediction of the standard competitive labour market model, at least for the range of NLW increases considered here.
The finding of a small but statistically significant reduction in average hours in high-bite sectors is a more nuanced result, and is consistent with a growing strand of the literature suggesting that firms respond to minimum wage increases through a range of margins beyond simple headcount, including hours, non-wage benefits, prices and productivity (Riley and Bondibene, 2017; Low Pay Commission, 2022). This is important for policy interpretation: a finding of ‘no employment effect’ based on headcount alone may understate the true labour market adjustment to a wage floor increase if firms are instead reducing hours or other forms of compensation. Future work extending this analysis should incorporate total labour input, that is hours multiplied by headcount, as an additional outcome, alongside available measures of non-wage benefits where data allow.
The results are broadly consistent with monopsony-based explanations of the low-wage labour market (Manning, 2021), in which employers possess some wage-setting power, for example because of search frictions, limited local job opportunities, or the non-pecuniary costs to workers of changing employer. Under such conditions, a minimum wage increase can raise both wages and employment, at least up to the point at which the wage floor approaches the competitive equilibrium wage. Sectors such as retail, hospitality and social care, which combine geographically dispersed employers, relatively low job mobility, and workers with fewer outside options, plausibly exhibit the features associated with monopsonistic wage-setting, which may help explain the pattern of results found here and in comparable UK studies.
Several limitations should be acknowledged. This analysis uses illustrative, stylised data constructed to be broadly consistent with published aggregate series rather than firm- or worker-level administrative data, and the resulting estimates should therefore be read as demonstrating an appropriate empirical approach rather than as definitive causal estimates. The relatively short, eight-quarter window limits the ability to detect employment effects that may emerge with a lag, for example as firms gradually adjust staffing levels, automate tasks, or exit the market over a longer horizon. The analysis is also conducted at sector level; effects may be considerably more pronounced for individual low-margin firms within high-bite sectors than the sector-level averages suggest, and small and medium-sized enterprises may face different constraints from larger firms in absorbing wage cost increases.
As the NLW continues to rise relative to median earnings, potentially exceeding levels for which minimum wage increases have historically been found to be employment-neutral, continued monitoring using this kind of difference-in-differences approach, ideally applied to disaggregated administrative data, will remain important for UK labour market policy.
This paper has reviewed the theoretical and empirical literature on minimum wage employment effects and applied a difference-in-differences framework, using illustrative UK-consistent data, to examine the impact of the National Living Wage on employment, hours and pay in low-paying sectors. Consistent with much of the recent international and UK evidence, no statistically significant negative effect of the NLW on aggregate employment was identified, while a strong, significant positive effect on hourly pay for low-paid workers was found, alongside a modest but significant reduction in average hours in more affected sectors. These findings are more consistent with monopsony-based models of the low-wage labour market than with the standard competitive prediction of substantial job losses, and support the view that, within the range of increases implemented so far, the NLW has succeeded in raising pay for low-paid workers without generating large-scale disemployment. As the wage floor continues to rise relative to median earnings, however, this relationship cannot be assumed to hold indefinitely, and rigorous, continually updated empirical monitoring, ideally using disaggregated administrative data, will remain essential to UK minimum wage policy.
Need a Model Research Paper Written to Your Exact Brief?
Our 350+ UK-qualified writers deliver referenced model documents from £15 per 250 words, with free plagiarism and AI-detection reports.
You May Also Like
All work is written by human writers. 100% AI free, guaranteed.
100% money back guarantee if you find plagiarism in our work.
COMPANY DETAILS
