Content uploaded by Max Collett
Author content
All content in this area was uploaded by Max Collett on Jul 09, 2024
Content may be subject to copyright.
2
Abstract
In discussions of climate change mitigation, working time reduction has emerged as a
potential win-win strategy to reduce emissions and enhance wellbeing. This dissertation
examines the potential impacts of working time reduction, in the form of the four-day week
with no loss of pay, on emissions in the UK. An economic model is developed to identify the
key channels linking working hours and emissions, which are then empirically investigated
via interviews with workers and managers that have adopted the four-day week. I find that
the four-day week would likely increase emissions in the short term by promoting carbon-
intensive consumption habits. Over longer timespans, emissions reductions from the four-day
week are possible, but depend on future productivity trends. I also show that the UK’s current
balance between income and leisure is misaligned towards the former, presenting significant
opportunity to realise environmental and wellbeing benefits via other forms of working time
reduction.
150 words
3
Acknowledgements
This dissertation would not have been possible without the extraordinary assistance afforded
to me by a range of people.
I owe a tremendous debt of gratitude to my supervisors, Prof. Sam Fankhauser and Dr.
Emilien Ravigné. It is only thanks to the thoughtful advice and positive energy of these
brilliant men that I have been able to produce a dissertation that I am proud of. Special
mention must go to Emilien, whose patience, encouragement, and generosity of spirit were
key sources of support during my crash course in economics. Of course, despite their best
guidance, I alone am responsible for any shortcomings in this work.
Thank you to my family, and particularly my parents, for their unwavering love and support.
I would not be the person I am, nor would I be here studying at Oxford, were it not for my
amazing mother and father. Without them, none of this would be possible.
I would also like to express thanks to the workers and managers who contributed their
valuable time and effort to this project. The opportunity to speak with such pioneering and
interesting people made the research process all the more enjoyable.
Thank you as well to my incredible friends, both inside and outside of Oxford, for their
comments on this work. More importantly, their camaraderie, good humour, and kindness
have played no small part in maintaining my sanity and making this year one of the most
enjoyable of my life to date.
Lastly, I am grateful for the various forms of support, academic or otherwise, offered by the
ECM course directors, other teaching staff within the SoGE, cleaners and support staff,
Hinksey Lake, University Parks, and Najar’s Place.
4
Table of Contents
Abstract ...................................................................................................................................... 2
Acknowledgements .................................................................................................................... 3
List of figures ............................................................................................................................ 6
List of tables .............................................................................................................................. 6
Abbreviations ............................................................................................................................. 7
Section 1: Introduction ............................................................................................................. 8
Section 2: Literature review .................................................................................................... 10
Section 2.1: Trends in working hours ............................................................................................10
Section 2.2: Introducing WTR and the post-growth framework ................................................11
Section 2.3: Environmental impacts of WTR ...............................................................................12
Section 2.3.1: Theoretical foundations .........................................................................................................12
Section 2.3.2: Empirical evidence ................................................................................................................13
Section 2.3.3: Reported environmental impacts of the recent 4-Day Week Pilot programmes ...................19
Section 2.4: Situating this dissertation ..........................................................................................20
Section 3: Analytical framework ............................................................................................ 21
Section 3.1: Modelling approach ....................................................................................................21
Section 3.2: Introducing the model ................................................................................................22
Section 3.2.1: Production side ......................................................................................................................22
Section 3.2.2: Consumption side ..................................................................................................................23
Section 3.2.3: Emissions ...............................................................................................................................24
Section 3.3: Equilibrium modelling ...............................................................................................25
Section 3.3.1: Setting up equilibrium conditions .........................................................................................25
Section 3.3.2: Finding equilibrium ...............................................................................................................26
Section 3.4: Theoretical analysis ....................................................................................................29
Section 3.4.1: Identifying the key channels ..................................................................................................29
Section 3.4.2: General equilibrium analysis .................................................................................................30
Section 3.5: Model findings .............................................................................................................37
Section 4: Qualitative methods ............................................................................................... 38
Section 4.1: Qualitative approach ..................................................................................................38
Section 4.2: Interview methodology ...............................................................................................38
Section 4.2.1: Participants ............................................................................................................................38
Section 4.2.2: Interview process ...................................................................................................................40
Section 4.2.3: Data analysis ..........................................................................................................................41
Section 5: Interview results .................................................................................................... 41
Section 5.1: Production ...................................................................................................................41
Section 5.1.1: 4DW format ...........................................................................................................................41
Section 5.1.2: Productivity and output .........................................................................................................42
Section 5.1.3: Energy and transport .............................................................................................................45
Section 5.1.4: Capital and investment ..........................................................................................................46
Section 5.1.5: Staffing ..................................................................................................................................46
Section 5.1.6: Carbon footprint ....................................................................................................................46
Section 5.2: Consumption ...............................................................................................................46
Section 5.2.1: 4DW format ...........................................................................................................................46
5
Section 5.2.2: Consumption and spending ...................................................................................................47
Section 5.2.3: Leisure travel .........................................................................................................................50
Section 5.2.4: Time use changes ..................................................................................................................50
Section 5.2.5: Trade-off between income and leisure time ..........................................................................52
Section 5.2.6: Day off preferences ...............................................................................................................54
Section 5.2.7: Carbon footprint ....................................................................................................................54
Section 6: Discussion .............................................................................................................. 55
Section 6.1: Key findings – does the 4DW decrease emissions? ..................................................55
Section 6.1.1: Interpretation of interview results using the analytical framework .......................................55
Section 6.1.2: Looking beyond the analytical framework ............................................................................59
Section 6.2: Policy discussion .........................................................................................................62
Section 6.3: Limitations and future directions .............................................................................64
Section 7: Conclusion ............................................................................................................. 65
References ............................................................................................................................... 66
Appendix A: Expressions for labour demand !" and capital # .......................................... 72
Appendix B: Labour supply function ..................................................................................... 74
Appendix C: Partial derivatives with respect to productivity ($) .......................................... 77
Appendix D: Proofs for Eqs. 33 and 34 ................................................................................. 78
6
List of figures
Figure 1
Average annual working hours per worker in the UK, 1950-2023
Figure 2
Schematic of relationship between working hours and emissions
Figure 3
Relationship between productivity and wage rate
Figure 4
Diagram of four possible scenarios following WTR
Figure 5
Impact of the 4DW on productivity
Figure 6
Impact of the 4DW on workers’ aggregate spending
Figure 7
Impact of the 4DW on workers’ time use patterns
List of tables
Table 1
Estimates of working hours elasticities of environmental impacts
Table 2
Theoretical responses of different variables to WTR
Table 3
Impact of the 4DW on composition of workers’ spending
Table 4
Empirically-determined effects of the 4DW on different variables
7
Abbreviations
BAU
Business as usual
DfT
Department for Transport
DEFRA
Department for Environment, Food & Rural Affairs
DESNZ
Department for Energy Security & Net Zero
EU
European Union
GDP
Gross domestic product
IPCC
Intergovernmental Panel on Climate Change
OECD
Organisation for Economic Co-operation and Development
ONS
Office for National Statistics
SECO
Switzerland State Secretariat for Economic Affairs
WTR
Working time reduction
8
Section 1: Introduction
Drastic reductions in greenhouse gas emissions are urgently needed to stabilise increasing
global temperatures and avert catastrophic climate change scenarios (IPCC, 2022). Given that
the bulk of historical responsibility for climate breakdown lies squarely with the Global
North, it is in the developed world that the low carbon transition ought to proceed first and
fastest (Hickel, 2020a). To this end, beyond the technological advances that dominate climate
change response discourses, social and economic innovations are essential to realise deeper
cuts to emissions.
At the same time, workers of many industrialised nations are confronted by the bleak reality
that, despite prodigious productivity gains, working hours have experienced only modest
declines over the last century (Huberman and Minns, 2007). Productivity growth has resulted
in soaring output, and environmental impacts, rather than increased time affluence, thwarting
John Maynard Keynes’ famous vision of a fifteen-hour working week for his grandchildren
(Keynes, 1930; Schor, 2005). Given the prevalence of overemployment1, and that longer
working hours are known to be associated with adverse health outcomes and lower life
satisfaction, there is a strong case to be made that shorter working hours may result in
substantial benefits to societal wellbeing (Alesina et al., 2005; Angrave and Charlwood,
2015; Wong et al., 2019).
Working time reduction (WTR) in the Global North has emerged as a potential means of
simultaneously addressing these issues. Trade unions have fought for shorter working hours
for decades, and have generally done so in promotion of either full employment or worker
wellbeing (De Spiegelaere and Piasna, 2017; Oh et al., 2012). Recent research, however,
suggests that WTR may also yield environmental benefits by decoupling GDP growth from
productivity growth, thus scaling down consumption, and affording people the time affluence
to pursue less resource-intensive lifestyles (Knight et al., 2013; Schor, 2010, 2005).
Advocates therefore claim that WTR offers a triple dividend to society: enhanced wellbeing,
lower unemployment, and reduced environmental pressures (Fitzgerald et al., 2018).
1 Overemployment occurs when a person would prefer to work fewer hours than they currently do, with no
change to their hourly wage rate.
9
In the 2020s, WTR, and particularly the four-day week with no loss of pay (4DW)2, has
received a crescendo of support in the UK and other developed countries from a chorus of
politicians, activists, economists, social scientists, and the public at large, even forming part
of the UK Labour Party’s policy agenda at the 2019 general election (The Labour Party,
2019). As such, experimentation with WTR has escalated rapidly: following the wave of
workplace innovation during the COVID-19 pandemic, 4DW pilots have sprung up across
the Global North. The success of such trials has generated prolific media coverage in support
of the concept (Godwin, 2023; Read et al., 2023). In 2023, snowballing political support has
led to parliamentary bills or recommendations endorsing the 4DW concept in Australia, the
UK, and the US. These developments have thus propelled the 4DW from the radical fringe
into mainstream policy debates, creating impetus for further inquiry into the concept and its
resultant impacts.
Consequently, this dissertation aims to investigate the potential for WTR, in the form of the
4DW, to reduce carbon emissions in the UK.
To this end, the following research questions are posed:
RQ1.
What are the key channels that determine the relationship between working
hours and emissions?
RQ2.
What are the potential effects of adopting a 4DW on carbon emissions in the
UK?
RQ3.
How might policy be crafted to promote emissions reductions via the 4DW
and/or other forms of WTR?
Each of these questions builds on the preceding one, and are therefore addressed in this order
throughout this dissertation. To address RQ1, I develop and analyse an economic model to
illuminate the key channels by which WTR is likely to influence emissions. Qualitative
research interviews are used to confirm these channels. Interview data are then analysed
2 In this dissertation, “4DW” refers to the four-day week with no loss of pay, unless otherwise specified.
10
using the model to address RQ2. Lastly, the findings of the first two research questions
inform a discussion of policy implications that addresses RQ3.
Following this introduction, Section 2 reviews the literature on the relationship between
working hours and environmental impacts. The economic model linking working hours and
emissions is presented and analysed in Section 3. Section 4 describes the qualitative
interview methodology employed to empirically assess the impact of the 4DW on emissions.
Interview results are then described in Section 5. Section 6 elucidates key findings and
discusses them in relation to the literature before describing potential implications for
policymaking, and is followed by conclusions in Section 7.
Section 2: Literature review
To begin this literature review, I briefly describe historical trends in working hours in the
UK. I then explicate the basis of the argument for WTR, with particular attention afforded to
the post-growth literature from which many such strands of argument originate, before
turning to critically review the literature on the environmental impacts of WTR. Finally, I
situate this dissertation in relation to existing literature by identifying the knowledge gaps
that this research seeks to address.
Section 2.1: Trends in working hours
In the UK, as in most industrialised countries, average working hours have fallen markedly
over the last century and a half (Huberman and Minns, 2007). This decline was particularly
steep through the interwar and post-war periods, but has stabilised since the mid-1970s (Fig.
1).
This presents a paradox of sorts: UK workers are working just 10% fewer hours than 50 years
ago despite productivity having more than doubled in that time (ONS, 2022; The Conference
Board, 2023). Coupled with the UK’s socio-political addiction to economic growth, two
major historical trends can explain this fact. First, Thatcher’s labour market deregulation in
the 1980s severely diminished the power of trade unions, leaving workers unable to argue the
case for shorter hours (Alesina et al., 2005; Mayhew, 1991). Second, productivity in the UK
has experienced an unprecedented slowdown since the Global Financial Crisis (Crafts and
11
Mills, 2020), meaning that economic growth can only be maintained by working longer
hours.
Working hours in the UK are generally on par with, or slightly less than, those in other
comparable economies. In 2016, the average UK worker toiled for 1515 hours, placing the
country nineteenth highest among twenty-six European OECD nations in terms of annual
working hours (Ward et al., 2018).
Figure 1: Average annual working hours per worker in the UK, 1950-2023. Source: The
Conference Board, 2023.
Section 2.2: Introducing WTR and the post-growth framework
Campaigns for WTR have traditionally been led by trade union movements, and have
predominantly argued the case from the perspective of reducing unemployment and
enhancing worker wellbeing (Alesina et al., 2005; Oh et al., 2012).
The environmental argument for WTR is somewhat newer, and has been advanced mostly by
degrowth, post-growth, and ecological economists (Costanza et al., 2013; Jackson, 2016;
Kallis et al., 2012; Victor, 2008). These perspectives emphasise that continual economic
0
500
1000
1500
2000
2500
1950 1960 1970 1980 1990 2000 2010 2020
Average annual working hours per worker
Year
12
growth in developed countries creates significant environmental and social harms, and
espouse a planned reduction in the production and consumption of goods that generate such
harms. This framework differs from the economic orthodoxy on environmental problems,
which generally seeks to reduce the resource intensity of economic activity while still
promoting continual economic growth – an approach termed “green growth” (Bowen and
Hepburn, 2014; Hallegatte et al., 2012; OECD, 2011).
WTR constitutes a key pillar in the post-growth framework (Hickel, 2020b; Jackson, 2016;
Schor, 2010). Under degrowth scenarios, working hours are reduced to avoid increasing
unemployment, and hence ensure post-growth’s social acceptability, amid a deliberate
slowdown in economic activity (Cieplinski et al., 2021b; Victor, 2008). Aside from its role as
a means of mitigating potential risks of a post-growth economic plateau, WTR itself may also
slow economic growth as, ceteris paribus, shorter working hours implies reduced production
and incomes, although the desirability of this depends on one’s view of the post-growth
argument (Booth and Schiantarelli, 1987; Cieplinski et al., 2021b; Knight et al., 2013; Schor,
2005).
Section 2.3: Environmental impacts of WTR
Section 2.3.1: Theoretical foundations
Theoretical arguments linking WTR to reduced environmental pressures first emerged in the
early 1990s. Such arguments are generally grounded in critiques of economic growth, and
frame WTR as a means of reducing consumption and transitioning to a more sustainable
economic model (Gorz, 1994; Hayden, 1999; O’Hara, 1993; Schor, 1995, 1992, 1991).
The literature identifies two mechanisms linking working hours and environmental impacts:
the scale effect and the composition effect (Hayden and Shandra, 2009; Knight et al., 2013).
The scale effect refers to the impact of average working hours on emissions via GDP: ceteris
paribus, working longer hours results in higher output, incomes and consumption, which
increases emissions. Working hours are thus understood to influence the scale of economic
activity, and therefore its attendant environmental impacts.
13
The composition effect refers to the effect of working hours on emissions net of GDP. This
effect describes the relationship between working hours and the carbon intensity of
consumption patterns – that is to say, on the composition of economic activity. It is perhaps
best understood at the household level as the effect of time affluence on the composition, and
hence carbon intensity, of time use and consumption.
Section 2.3.2: Empirical evidence
An array of studies have sought to assess the environmental impacts of WTR by various
methodologies. The literature broadly agrees that shorter working hours are associated with
reduced environmental impacts, and that this relationship is, for the most part, driven by scale
effects (Cieplinski et al., 2021b; Fitzgerald et al., 2018; Knight et al., 2013; Nässén and
Larsson, 2015). However, two key disagreements persist in the scholarship: first, estimates of
the working hours elasticities of environmental impacts3 vary dramatically between studies;
second, the direction and strength of the composition effect remain contested. Moreover, this
literature is broadly constrained by crucial methodological limitations and contains several
key gaps. The remainder of Section 2.3.2 is structured to cover each of these points.
Working hours elasticities of environmental impacts
The first attempt to empirically assess the macrostructural relationship between working
hours and environmental impacts was Schor’s (2005) exploratory analysis of 18 OECD
countries, where she found a significant, positive relationship between average working hours
and nations’ ecological footprints.
Subsequent econometric analyses have tended to confirm this positive relationship, yet have
produced varying estimates of the working hours elasticities of environmental impacts, as
displayed in Table 1.
The variation between these estimates can be explained by several factors. First, elasticity
estimates pertaining to territorial emissions appear to be significantly lower than those
measured using consumption-based metrics (Fitzgerald, 2022; Fitzgerald et al., 2018; Knight
3 The working hours elasticity of an environmental indicator expresses the percentage change in the indicator
that corresponds to a 1% change in working hours. For instance, if the working hours elasticity of carbon
emissions was 0.40, we would expect carbon emissions to increase (decrease) by 0.40% for every 1% increase
(decrease) in working hours.
14
et al., 2013; Mallinson and Cheng, 2021). This may be due to the fact that developed
countries tend to be net importers of emissions (Peters and Hertwich, 2008; Wilting and
Vringer, 2010), implying that analyses of consumption-based and territorial indicators may
produce divergent results.
Second, estimates based on household-level data are substantially lower than those calculated
using macro-level data (Fremstad et al., 2019; Nässén and Larsson, 2015). Fremstad et al.
(2019) offer a twofold explanation for this difference: first, cross-country studies ignore other
policy determinants of carbon emissions, leading to an upward bias in elasticity estimates;
second, countries may be better than households at reducing emissions via shorter working
hours due to social multiplier effects.
Some analyses have, however, found the relationship between working hours and emissions
to be negative over certain time periods and in certain geographies (Shao and Rodríguez-
Labajos, 2016; Shao and Shen, 2017). However, the results of these studies have been
roundly questioned based on the shakiness of their methodologies (e.g. see Antal et al.
(2020)).
15
Table 1: Estimates for working hours elasticities of environmental impacts from the
literature, adapted from a similar table in Shao and Shen (2017). Studies are listed in
chronological order.
Study
Indicator
Elasticity
estimate
Methodology
Data
Rosnick and
Weisbrot (2006)
Energy
consumption
1.33
Multivariate
regression, cross-
sectional
48 countries,
2003-2005
Hayden and
Shandra (2009)
Ecological
footprint
1.20
Multivariate
regression, cross-
sectional
45 countries,
2000
Knight et al.
(2013)
Ecological
footprint
1.37
First-difference
panel regression
27 high-income
OECD countries,
1970-2007
Carbon footprint
1.30
Carbon
emissions
0.50
Nässén and
Larsson (2015)
Energy
consumption
0.76
Micro-data analysis;
multivariate
regression
Swedish
households, 2006
Consumption
emissions
0.80
Fitzgerald et al.
(2018)
Carbon
emissions
0.67
Prais-Winsten
regression
US states, 2007-
2013
Fremstad et al.
(2019)
Consumption
emissions
0.32
Micro-data analysis;
multivariate
regression
US households,
2012-2014
Mallinson and
Cheng (2021)
Carbon
emissions
0.54
Prais-Winsten
regression
US states, 2007-
2016
Fitzgerald
(2022)
Carbon
emissions
0.99
Prais-Winsten
regression
US states, 2005-
2015
Working hours, the macroeconomy, and environmental impacts
Most econometric analyses listed in Table 1 agree that the majority, if not the entirety, of
environmental indicators’ responses to changes in working hours arise via the scale effect.
16
For example, Nässén and Larsson (2015) estimate that, ceteris paribus, a household’s
consumption emissions would fall by 0.82% with a 1% decrease in income, but rise by 0.02%
with a 1% decrease in working hours, supporting the theory that WTR mostly acts on
emissions by reducing consumption via the scale effect.
The literature thus suggests that to the extent that shorter working hours reduce emissions,
they do so mostly via slowing economic growth. It is therefore useful to examine the
interactions between WTR and other economic channels that contribute to GDP growth.
Assuming stable employment, GDP can be expressed as the product of average working
hours per worker and labour productivity – understood as output per person-hour of work.
Most studies generally find that both of these variables have significant, positive relationships
to environmental impacts (Fitzgerald, 2022; Hayden and Shandra, 2009; Knight et al., 2013).
However, it is also true that WTR itself may affect labour productivity (Golden, 2012). As
such, I now turn to examine the relationship between WTR, labour productivity and
environmental impacts.
The 4DW is understood as a form of productivity-led WTR, defined herein as the reduction
of working hours by an amount proportional to a more or less contemporaneous increase in
labour productivity (Cieplinski et al., 2021b). WTR of this kind is distinct from alternative
scenarios in which workers’ incomes decline proportionally to the reduction in working hours
or per worker output, thereby effectively holding the hourly wage rate stable or only
increasing it slightly.
While increasing labour productivity via technological progress or other innovations can
create the conditions for productivity-led WTR, the reduction of working hours in itself may
generate endogenous productivity gains arising from healthier, happier workers (Collewet
and Sauermann, 2017; Golden, 2012; Hanna et al., 2005). However, despite the success of
recent 4DW trials (Lewis et al., 2023; Schor et al., 2022), it cannot be assumed that a 4DW
will automatically lead to a proportional increase in labour productivity in all settings. Firstly,
to avoid declining output under a 4DW, productivity must increase by 25% to compensate for
the 20% reduction in working hours – a tall ask given the UK’s unprecedented productivity
slowdown (Crafts and Mills, 2020). 4DW advocates may reply that WTR itself could unlock
productivity gains, as discussed above. But, and secondly, 4DW trials have heretofore
focussed on white collar sectors in which productivity gains may be realised more easily, for
17
example via cutting down on meetings, than in service-based industries or manufacturing
(Lewis et al., 2023). In summary, the literature suggests that a 4DW is likely to have a
positive impact on labour productivity, but the size of this impact, and its resultant capacity to
maintain stable output with shorter working hours, requires further exploration.
There are, however, trade-offs associated with productivity growth in the WTR context.
Setting aside environmental concerns, the appeal of productivity-led WTR is straightforward:
if working hours are reduced proportionally with productivity gains, then output remains
stable, thus neutralising the trade-off between leisure time, income, employment, and profits.
However, the equation is complicated when one adds environmental objectives to the mix. As
discussed, if WTR is to reduce emissions, the scale effect is likely to be the dominant
mechanism at play. But, in the case of productivity-led WTR, GDP, incomes, and
employment theoretically remain stable, thus weakening the scale effect. A trade-off thus
emerges: productivity growth under WTR protects wages, employment, and profits, but
reduces the potential environmental benefits.
It should be noted, however, that by translating productivity gains into leisure time rather
than economic growth, productivity-led WTR does reduce emissions relative to the
counterfactual BAU (business as usual) scenario (Cieplinski et al., 2021b). In developed
countries, productivity gains have historically been translated mostly into higher output and
consumption, rather than extra leisure time (Schor, 2005). It is this tendency to prioritise
economic expansion that has driven the “work-and-spend cycle” described by Schor (1992)
and led to continuous consumption and emissions growth.
The effects of WTR on employment are complex, contested, and almost certainly depend on
a variety of factors and assumptions concerning productivity, wage responses, labour
bargaining, and policy settings (Booth and Schiantarelli, 1987; Cieplinski et al., 2021b; De
Spiegelaere and Piasna, 2017; Kapteyn et al., 2004; Roos, 1935). A full account of the
literature on this subject is omitted here because insofar as a 4DW in the UK will affect
emissions, it will not do so via employment effects, as it is assumed that stable employment
will be a requisite condition of any 4DW policy rollout.
18
Composition effect debate
The second major source of disagreement in the literature on the environmental impacts of
WTR is the size and direction of composition effects. Essentially, it remains unclear whether
consumption becomes more or less carbon intensive when working hours are reduced,
independent of income (at the household level) and GDP (at the societal level).
On the one hand, WTR produces time affluence, which may allow people to engage in
slower, less carbon-intensive lifestyles (Jalas, 2002). On the other hand, WTR could lead to
rebound effects if people dedicate their newfound leisure time to carbon-intensive activities.
The empirical evidence on the subject remains inconclusive. Most macrostructural analyses
find evidence of a positive, significant composition effect: Hayden and Shandra (2009)
estimate that a 1% reduction in working hours is associated with a 0.59% reduction in
ecological footprint, independent of GDP, and other studies have yielded similar estimates
for other environmental indicators (Fitzgerald, 2022; Fitzgerald et al., 2018; Mallinson and
Cheng, 2021). Conversely, Knight et al. (2013) find composition effects to be insignificant in
analyses of national carbon footprints and carbon emissions. The household data is equally
ambiguous. Devetter and Rousseau (2011) find that net of income, longer working hours are
associated with higher consumption of environmentally-intensive goods. Yet, other studies
find the effect of time affluence on consumption emissions, net of income, to be close to zero
(Fremstad et al., 2019; Nässén and Larsson, 2015).
It should be noted that estimates of composition effects derived from macro-level data tend to
be larger than those calculated from household data, which suggests that a societal approach
to WTR is more effective at shifting the carbon intensity of consumption patterns than a
household-level approach.
Furthermore, Hanbury et al. (2019) interviewed 17 Swiss employees that had recently made
self-determined reductions to their working hours and found that those who did so seeking
more recreational time risked increasing the carbon intensity of their lifestyles.
19
In the case of the 4DW, any impact on emissions is likely to arise via composition effects
since incomes remain unchanged. As such, given the lack of consensus regarding these
effects, the overall impact of the 4DW on emissions is not clear a priori.
Figure 2: Schematic showing the main channels that are presumed to impact carbon
emissions during the WTR transition. Single-headed and double-headed arrows indicate
unidirectional and bidirectional causal relationships, respectively. Although WTR, if
implemented at any appreciable scale, may have flow-on effects to other businesses and
people that still follow standard five-day full-time schedules, this schematic only represents
the effects that occur in workplaces and workers that adopt WTR. Source: Author.
Section 2.3.3: Reported environmental impacts of the recent 4-Day Week Pilot programmes
To date, two major rounds of 4-Day Week Pilot programmes have been conducted by the
organisation 4-Day Week Global. These pilot programmes involved dozens of organisations
across multiple countries trialling the 4DW on a six-month basis.
20
A brief evaluation of environmental impacts is contained in each of the post-trial reports
(Lewis et al., 2023; Schor et al., 2022). In terms of travel, large reductions in commuting
were observed, but leisure travel increased significantly. A small increase in household
energy use is also reported for the first trial. Together, these results suggest that household
consumption emissions may have increased due to the 4DW, but, as detailed below, their
robustness is constrained by poor research design.
These assessments rely on extremely problematic methodologies. For example, the frequency
of international travel is assessed by surveying the “number of round-trip international flights
taken within the last four weeks”, yet the trial lasted six months with employee surveys
completed at the beginning, middle and endpoints. This implies that any international travel
occurring in months 1-2 and 4-5 of the trial, spanning two-thirds of the trial’s duration, was
not counted. Moreover, seasonal variation in energy use and travel patterns was not
considered. As such, there remains no decisive analysis of the environmental impacts of these
trials.
Section 2.4: Situating this dissertation
This dissertation addresses several key gaps in the literature. First, as outlined above, the
UK’s recent 4-Day Week Pilot programme presents a valuable quasi-experiment that has
hitherto been underutilised in terms of understanding the environmental impacts of WTR. By
engaging with participants from these trials, this dissertation will be the first attempt to
empirically assess the emissions impact of the 4DW in an academic setting.
Second, the econometric approaches that dominate the literature, while useful in revealing
correlations, do not provide evidence of any causative effects linking WTR to reduced
environmental pressures. By interviewing managers and workers that transitioned to the
4DW, I am able to explicitly identify cause and effect.
Moreover, with the exception of Hanbury et al. (2019), no attempts have been made to
qualitatively understand the actual mechanisms shaping the relationship between WTR and
emissions. I address this paucity of understanding by employing qualitative methods to
examine the effects of the 4DW at both the individual and firm level.
21
Third, as explicated in Section 2.3.2, the size and direction of the composition effect is
ambiguous. The 4DW provides the perfect setting to analyse this effect, as employees
reduced their working hours without any loss of compensation. Any changes to their
consumption and time use patterns can therefore be attributed to the effects of changes in
time affluence. As such, this dissertation will shed new light on the size, direction, and
underlying behavioural explanation of the composition effect.
Section 3: Analytical framework
Section 3.1: Modelling approach
In this section, I adopt a neoclassical economic approach and develop an equation-based
model to illustrate the various channels by which WTR influences emissions. I then analyse
the model to elucidate key theoretical insights into the impacts of WTR, and finally identify
four possible scenarios that may emerge following changes to working hours and
productivity.
The shortcomings of neoclassical economics and equation-based modelling are well-
documented (Hausman, 1992; Keen, 2001). For brevity’s sake, I restrict my discussion to
three key criticisms and respond to them accordingly.
First, the neoclassical approach often entails making various simplifying assumptions about
the way markets, firms, and people behave (Hausman, 1992; Kirman, 1992; Sen, 1977;
Sensat and Constantine, 1975). For example, in my model, I ignore unemployment, use one
representative utility-maximising agent to model heterogenous consumer preferences, and
assume perfect markets. I justify this on the basis that the purpose of this exercise is to
identify the principal channels through which WTR influences emissions in the short-term; I
do not aim to conduct precise quantitative simulations.
Second, most neoclassical models, including my own, do not represent feedback loops
between the economy and the environment (Daly, 2015). I justify this approach on the basis
that, in the short-medium term at least, environmental phenomena are unlikely to affect the
relationship between working hours and emissions.
22
Third, equation-based modelling is often criticised as being more concerned with
mathematical gymnastics than explaining real-world phenomena (Georgescu-Roegen, 1971;
Hausman, 1992). In this dissertation, however, I combine the modelling approach with
qualitative, empirical investigation, thus bridging the gap between theoretical and empirical
methods of inquiry. In this sense, the model is not deployed in a vacuum of theoretical
impunity, but is tested against and complemented by empirical evidence.
Section 3.2: Introducing the model
The model developed in this section consists of a production side, on which firms combine
capital and labour to produce one good, and a consumption side, where people derive utility
from two sources – consumption and leisure. Emissions are then modelled as a function of
consumption, leisure, and the carbon intensities of these variables.
Section 3.2.1: Production side
Total output value %, is expressed using a standard Cobb-Douglas production function – as a
combination of labour input &, understood as person-hours of labour, and capital input ',
measured as the value of the capital stock.
Eq. 1
% ( )'!&"#!
) is a residual term and refers to total-factor productivity. The constant * denotes the output
elasticity of capital, also approximated as the capital share of income4, and takes values + ,
* , -.
Firms’ profits . are calculated as turnover minus costs; that is, output, with price normalised
at 1, less costs of labour, paid the hourly wage rate /, and capital paid interest 0.
Eq. 2
. ( 1)'!&"#! 2/& 20'
4 The UK’s labour share of income is around 60% (ONS, 2023c), implying a value of ! of 0.4.
23
Section 3.2.2: Consumption side
Households derive utility 3 from two activities: consumption 4 and leisure 5 (for fun). This
formulation is common in studies of labour markets (Cutanda and Sanchis-Llopis, 2019; de
Fraja, 1996; Marimon and Zilibotti, 2000).
Eq. 3
3 ( 364758
Consumption is measured in monetary terms and refers to aggregate household spending. It is
assumed that households spend all their income and do not save or borrow, which is a
relatively common assumption in the literature (Marimon and Zilibotti, 2000); thus:
Eq. 4
4 ( /&
Leisure is measured in temporal terms and defined by the assumption that people’s total time
endowment 9 is split between leisure and labour:
Eq. 5
5 ( 9 2 &
Assume workers exhibit Cobb-Douglas preferences regarding consumption and leisure:
Eq. 6
364758( 4$5"#$
The parameter : describes households’ preference for consumption over leisure and takes the
values + , : , -. A high value of : (i.e. : ; +<=) implies that people derive more utility
from consumption rather than leisure, and the opposite is true for low values of :.
24
Section 3.2.3: Emissions
Assume that carbon emissions arise from both consumption and leisure. Consumption
emissions refers to emissions embodied in the production of goods that are consumed.
Leisure emissions refers to emissions that arise from activities pursued during one’s leisure
time. Total emissions from the consumption side of the economy > can thus be represented
as the sum of these two variables:
Eq. 7
> ( ?%4 @?&5
?% refers to the carbon intensity of consumption, measured in tCO2e/£, and ?& to that of
leisure time, measured in tCO2e/hr.
The separation of consumption and leisure is necessary to link changes in working hours and
consumption to changes in emissions via the equilibrium approach described in Section 3.3.
However, this formulation comes with the caveat that it may be empirically challenging to
distinguish between the two channels, given that some consumption occurs during one’s
leisure time. This concern can be addressed by defining the separation and calibrating ?&
such that the two channels are mutually exclusive, thus avoiding double-counting.
This leisure emissions intensity term ?& may be better understood by expressing it the form:
Eq. 8
?&(>&
4&A4&
5
>& represents emissions from leisure activities and 4& represents leisure consumption. The
emissions intensity of leisure time is therefore calculated as the emissions intensity of leisure
consumption multiplied by the consumption intensity of leisure time.
25
Section 3.3: Equilibrium modelling
In this section, I first derive the equilibrium conditions under which labour demand (from
firms) is equal to the labour supply (from households) given profit and utility maximisation. I
then solve the model to obtain expressions for the different channels at equilibrium.
Section 3.3.1: Setting up equilibrium conditions
Production side
Firms maximise profits by selecting optimal quantities of labour and capital inputs. Under the
first order conditions '(
') ( + and '(
'* ( +, expressions for labour demand &+ and capital '
under profit-maximising conditions are given by5:
Eq. 9
&+( B)6-2*8C"
!'/#"
!
Eq. 10
' ( D "
"#,&E0
*F"
!#"
Consumption side
Households maximise their utility by choosing optimal quantities of consumption and leisure,
constrained by the conditions defined in Eqs. 4 and 5. Using Lagrangian methods, it can be
shown that maximum utility is achieved under the condition6:
Eq. 11
G3
G4/ ( G3
G5
Note that in this equation, '-
'% and '-
'& are the partial derivatives of the utility function with
respect to consumption and leisure, respectively.
5 See Appendix A for working.
6 See Appendix B for working.
26
This result implies that workers will seek to set their working hours such that the marginal
utility of increasing consumption by £/ (by working one extra hour) is equal to the marginal
disutility of losing one hour of leisure time.
To obtain the labour supply function, I take the partial derivatives of the utility function
described by Eq. 6 with respect to consumption and leisure time, then substitute these
expressions into Eq. 11 and solve for &, giving7:
Eq. 12
&.(:9
The worker’s optimal working hours are thus determined solely by her preference for
consumption over leisure. Intuitively, valuing one’s leisure time highly will reduce the hours
one is willing to work, whilst a stronger preference for consumption over leisure time will
lead one to work longer hours to increase their income and, hence, consumption.
This formulation is simpler than other labour supply functions insofar as it does not depend
on wages, labour bargaining power, or any other labour market variables. This is acceptable
in the context of this study, as I assume that WTR does not occur as a response to shifting
wages, but from the desire of workers or employers for enhanced work-life balance.
Section 3.3.2: Finding equilibrium
In this section, I derive expressions for labour, capital, output, wages, consumption, and
leisure time at equilibrium. I assume international trade clears the goods market, so it is not
assumed that % ( 4.
7 See Appendix B for working.
27
Since labour supply &. is independent of wages and depends only on :, the equilibrium
quantity of labour &∗ is simply equal to the labour supply:
Eq. 13
&∗( &+( &.(:9
The equilibrium quantity of capital '∗ is determined by substituting this result into Eq. 10,
giving:
Eq. 14
'∗( ) "
"#!:9E0
*F"
!#"
The expression for output at equilibrium %∗ is obtained by substituting the above expressions
for &∗ and '∗ into the production function (Eq. 1), giving:
Eq. 15
%∗( ) "
"#!:9E0
*F!
"#!
Substituting these expressions for &∗ and '∗ into the labour demand expression (Eq. 9) gives
the equilibrium wage rate required to clear the labour market /∗:
Eq. 16
/∗( ) "
"#!6-2 *8E0
*F!
!#"
It is important to note here that, given constants * and 0, the equilibrium wage rate increases
with ) according to a power law, implying that higher productivity leads to higher wages
(Fig. 3).
28
Figure 3: Relationship between productivity and wage rate given constants * and r.
Given that 4∗( /∗&∗, the equilibrium expression for consumption is given by:
Eq. 17
4∗(:9)"
"#!6-2 *8E0
*F!
!#"
An expression for leisure time under equilibrium conditions can also be derived, given that
5∗( 9 2&∗.
Eq. 18
5∗(6-2 :89
Emissions at equilibrium >∗ can then be expressed by substituting expressions for /∗, &∗ and
5∗ into Eq. 7:
Eq. 19
>∗( ?%:9)"
"#!6-2 *8E0
*F!
!#" @ ?&6- 2:89
Wage rate
Productivity
29
Section 3.4: Theoretical analysis
In this section, I first highlight the key dynamics that are presumed to determine the impact
on emissions, according to the model. I then proceed to analyse the model using comparative
statics, where I examine the impacts of WTR and productivity gains on emissions and output.
Section 3.4.1: Identifying the key channels
Given that 5 ( 9 2 & and 4 ( /&, the expression for emissions (Eq. 7) can be written in
terms of /, & and the carbon intensity terms as:
Eq. 20
> ( ?%/&@ ?&692&8
It is clear from this equation that emissions impacts are essentially determined by three
channels: the allocation of time between labour and leisure (& and 5), the carbon intensities
of consumption (?%) and leisure (?&), and consumption (4). Consumption is written here as
/& (see Eq. 4) and is determined by wages, and hence productivity (see Eq. 16). The effects
of WTR on these channels combine to determine the overall impact on emissions. I include
brief summaries of each of these three channels below and, following theoretical analysis,
investigate them empirically using qualitative methods (see Sections 4-6).
First, the emissions impact of reducing &, with no changes to the other variables, is not
immediately clear. Indeed, emissions will only fall under the condition '0
') ; +, which occurs
when:
Eq. 21
?&, /?%
This condition makes sense intuitively: the LHS (left-hand side) represents the emissions of
one hour of leisure, and the RHS represents the emissions from £/ of consumption – the
consumption of one hour’s worth of wages. If the emissions from consuming one hour’s
worth of wages are greater than those from one hour of leisure, then reducing & and
30
sacrificing some amount of income and consumption will reduce emissions. This will
generally be the case when / and/or ?% are large, and ?& is small.
Second, it is clear that reductions in the carbon intensity terms ?% and/or ?& will, ceteris
paribus, cause emissions to fall. Indeed, this scenario represents the observed trajectory of
the UK economy in recent years. The UK’s consumption emissions peaked in 2007 and fell
30.1% to 2019 (DEFRA, 2023). Meanwhile, average working hours and real wages did not
budge in this period (ONS, 2023a; The Conference Board, 2023). According to the model,
the observed decline in emissions must therefore be attributed to reductions in the carbon
intensity terms.
These carbon intensity terms may decrease via two principal mechanisms. First, consumer
preferences may shift to less carbon-intensive products. Alternatively, supply-side
decarbonisation would reduce the carbon intensity of consumption without any change to
consumer preferences.
Lastly, increases to the wage rate / via productivity growth will generate upwards pressure
on emissions by boosting consumption, and must be accompanied by decreases in & or
carbon intensity if emissions are to remain stable. With the 4DW for example, the wage rate
increases proportionally with the decrease in working hours such that incomes, and hence
consumption, do not change. Since leisure time 5 increases, it follows that, according to the
emissions model, emissions will increase unless ?% and/or ?& decreases.
Section 3.4.2: General equilibrium analysis
In this section, I first examine the impacts of WTR and productivity growth in isolation on
the economy and emissions via comparative statics, then extend the analysis to look at the
combined effects of simultaneous WTR and productivity growth.
Comparative statics
In this analysis, I assume that WTR occurs as a result of people’s desire for more leisure
time. Hence, I examine the effects of WTR via the shifts in equilibria following a change to :
(see Eq. 13).
31
Furthermore, since WTR is likely to generate productivity gains (Collewet and Sauermann,
2017; Golden, 2012), I also investigate the impacts of an increase in ).
The impacts of a reduction in : can be identified by inspecting the partial derivatives of
labour &∗, leisure time 5∗, consumption 4∗, capital '∗, and output %∗ with respect to ::
Eq. 22
G&∗
G: ( 9
Eq. 23
G5∗
G: ( 29
Eq. 24
G4∗
G: ( 9) "
"#!6-2 *8E0
*F!
!#"
Eq. 25
G'∗
G: ( 9) "
"#! E0
*F"
!#"
Eq. 26
G%∗
G: (9) "
"#! E0
*F!
"#!
It is clear that labour, consumption, capital, and output will all decrease following a reduction
in :, as their respective partial derivatives are positive for all possible values, whereas the
opposite is true for leisure time (Table 2).
32
Note that the equilibrium wage rate /∗ does not vary in response to :, as it depends on
productivity ()), not on people’s propensity to consume (see Eq. 16).
A shift from consumption towards leisure time will also occur following a reduction in :, as
'1!∗
#∗2
'$ , given in the below equation, is necessarily positive:
Eq. 27
GE4∗
5∗F
G: ()"
"#!6-2 *8E0
*F!
!#"
6-2 :83
Regarding emissions, the partial derivative of emissions at equilibrium >∗ with respect to : is
obtained from Eq. 19:
Eq. 28
G>∗
G: ( ?%9) "
"#!6-2 *8E0
*F!
!#" 2 ?&9
If '0∗
'$ is positive, then a reduction in :, symbolising WTR, will cause emissions to fall. '0∗
'$
will be positive when:
Eq. 29
?&, ?%)"
"#!6-2 *8E0
*F!
!#"
The LHS of this inequality represents the emissions associated with one hour of leisure time.
The RHS shows the carbon intensity of consumption (?%) multiplied by the equilibrium wage
rate given in Eq. 16, which together give the emissions from £/ of consumption. This
essentially represents the same condition identified by Eq. 21, but expressed in terms of the
parameters that define /∗.
33
Thus, according to the model, WTR will only reduce emissions if the emissions from
consuming one hour’s worth of wages are greater than the emissions from one hour of leisure
time, assuming no changes to other terms, since people are effectively trading the former for
the latter.
The effects on each of the different variables arising from a reduction in : are displayed in
Table 2.
Turning now to productivity, the effects of productivity growth on the various channels are
determined by their respective partial derivatives8 with respect to ). These impacts are
displayed in Table 2.
Since consumption increases with productivity in the model due to wage growth (see Eq. 17),
this would create upwards pressure on emissions:
Eq. 30
G>∗
G) ( ?%:9E0
*F!
!#" )!
"#!
Each term of Eq. 30 is necessarily positive, so '0∗
'4 is positive for all possible values of ),
which means that, ceteris paribus, any increase (decrease) in productivity causes an increase
(decrease) in emissions.
8 See Appendix C for partial derivatives.
34
Table 2: Responses of variables, ceteris paribus, to a reduction in :, symbolising WTR via
increased preference for leisure, and an increase in ), symbolising productivity gains.
Variable
Effect of reduction in
H
Effect of increase in
$
Labour &∗
Decrease
No change
Leisure time 5∗
Increase
No change
Consumption 4∗
Decrease
Increase
Ratio of consumption to
leisure time %∗
&∗
Decrease
Increase
Wage rate /∗
No change
Increase
Capital '∗
Decrease
Increase
Output/GDP %∗
Decrease
Increase
Emissions >∗
Ambiguous – will decrease if
"$# "%$!
!"#%& ' !()&
'*
#
#"!
Increase
As highlighted in Table 2, the effects of productivity gains are often contrary to those arising
from reductions in :. In the case of emissions, these effects occur because productivity gains
increase the wage rate, and hence consumption. Consequently, when WTR is accompanied
by productivity gains, as is the case with the 4DW, the impacts on emissions and other
variables cannot be predicted a priori, as they depend on the magnitude of the changes in )
and :.
Combined effects of productivity growth and WTR
In this section, I examine the impacts of changes to productivity ) and working hours (via
leisure preferences :) on the equilibrium values of output %∗ and emissions >∗ to define four
possible economic trajectories: conventional growth, green growth, degrowth, and policy
failure.
35
First, the expression for equilibrium output at time I is obtained from Eq. 15:
Eq. 31
%5∗( )5
"
"#!:59E0
*F!
"#!
Equilibrium emissions at time I is given by:
Eq. 32
>5
∗( ?%:59)5
"
"#!6-2 *8E0
*F!
!#" @ ?&6- 2:589
Suppose that : is modified by a factor J where + , J , - such that :56" ( J:5. Note that
this implies WTR, since &56"
∗( J&5
∗.
Productivity changes by a factor K, where K ; -. Productivity at time I @- is therefore
represented as )56" ( K)5
It can then be shown that output will fall (i.e. that %56"
∗, %5∗) under the condition9:
Eq. 33
J , K "
!#"
For example, supposing * ( +<L (as noted in Section 3.2.1), a 2.0% increase in productivity
(K ( -<+M) can be accompanied by a 3.2% reduction in working hours (J ( +<NOP) to hold
output stable. If working hours are reduced by more than 3.2%, then output would decline.
9 See Appendix D for working.
36
Similarly, it can be shown that emissions will fall (i.e. >56"
∗, >5
∗) under the condition10:
Eq. 34
?%)5
"
"#!6-2 *8E0
*F!
!#" Q- 2JK "
"#!R ; ?&6- 2J8
The LHS of Eq. 34 shows the multiplicative product of the carbon intensity per pound of
consumption (?%), the equilibrium wage rate (since /∗( ) (
()*6- 2*8E7
!F
*
*)(), and the factor
6-2 JK (
()*8. Hence, the LHS effectively represents the emissions associated with consuming
one hour’s worth of wages, modified by the factor 6-2 JK (
()*8, which implies that changes
to both consumer preferences (via J) and productivity (via K) influence consumption
emissions. The LHS can thus be understood as the difference in consumption emissions at
times I and I@-.
The RHS represents the emissions associated with one hour of leisure time (?&) multiplied by
the factor 6-2 J), or, in other words, the difference in leisure emissions at times I and I @-.
Note that leisure emissions depend only on variations in working hours (represented by J),
not on productivity changes.
Overall, this condition implies that, following changes to : and productivity ()), emissions
will fall if the decrease in consumption emissions (on the LHS) is greater than the increase in
leisure emissions (RHS), assuming no changes to the carbon intensity terms.
Inspecting Eq. 34, it can also be noted that deeper cuts to working hours (smaller J) will
increase both sides of the expression. Higher values of K, corresponding to larger productivity
gains, will decrease only the LHS. Large productivity gains thus create upwards pressure on
emissions, whereas the impact of reduced hours due to increased desire for leisure time is
more difficult to gauge without empirical calibration.
10 See Appendix D for working.
37
From the conditions described by Eqs. 33 and 34, I construct four possible scenarios of future
output and emissions trajectories, as shown in Figure 4. The scenarios are defined by two
inequalities: the relative changes in productivity ()) compared to households’ preference for
consumption (:), and the marginal emissions of leisure compared to those of consumption.
Output (%∗) decreases
Output (%∗) increases
Emissions
(>∗) increase
POLICY FAILURE
J , K "
!#"
and
!!""
"#$#$ % &' ()
&*
$
$#" ($ % +, "
"#$*- !%#$ % +'
CONVENTIONAL GROWTH
J ; K "
!#"
and
!!""
"#$#$ % &' ()
&*
$
$#" ($ % +, "
"#$*- !%#$ % +'
Emissions
(>∗) decrease
DEGROWTH
J , K "
!#"
and
!!""
"#$#$ % &' ()
&*
$
$#" ($ % +, "
"#$*. !%#$ % +'
GREEN GROWTH
J ; K "
!#"
and
!!""
"#$#$ % &' ()
&*
$
$#" ($ % +, "
"#$*. !%#$ % +'
Figure 4: Diagram depicting four possible trajectories of output (%∗) and emissions (>∗)
following changes to productivity ()) by factor K and households’ preference for
consumption (:) by factor J. The uppermost inequality in each of the four boxes determines
changes to output, while the lower one determines changes in emissions.
Section 3.5: Model findings
Three principal conclusions can be drawn from the modelling and analysis described in
Section 3.
38
First, the impact of WTR on emissions is ambiguous. It is not true that reducing working
hours will necessarily decrease emissions. Indeed, if leisure time is more carbon intensive
than consumption, then WTR may, ceteris paribus, increase emissions.
Second, the effects of productivity growth tend to work in opposition to the effects of shorter
working hours (see Table 2). Hence, when WTR and productivity growth occur in concert,
the economic and emissions impacts cannot be known a priori. Thus, whether WTR leads to
green growth, degrowth, conventional growth or policy failure is determined by the specifics
of the WTR programme (Fig. 4).
Third, the impact of WTR on emissions depends on several key channels, including
consumption and wages, driven by productivity growth, and the carbon intensities of
consumption and leisure. I employ qualitative research methods to empirically investigate
these channels in the context of the 4DW, as described in Section 4.
Section 4: Qualitative methods
Section 4.1: Qualitative approach
Empirical data were collected via qualitative research interviews. This methodology was
chosen with three principle aims in mind. First, the empirical approach complements the
theoretical inquiry in Section 3 by examining real-world phenomena, thus enabling richer
analysis of the subject at hand. Second, the interview process was designed to explicitly
estimate the impacts of the 4DW on emissions via the key channels identified in Sections 2
and 3. Third, qualitative methods were chosen to facilitate a deeper examination of the
underlying mechanisms responsible for observed effects – to explain how and why they
occur. Moreover, the ethical advantages of participatory research give value to the act of
knowledge production in itself (Cornwall and Jewkes, 1995).
Section 4.2: Interview methodology
Section 4.2.1: Participants
Twenty-four semi-standardised, individual interviews were conducted with people who work
at firms and organisations that have recently transitioned from a five-day week to the 4DW
with no loss of pay.
39
Workplaces that have adopted the 4DW were identified from the public list of companies that
took part in the 2022 UK 4-Day Week Pilot programme (Solomons et al., 2023), as well as
other online media. Participants were then recruited by writing to these organisations’ public
email addresses, and by contacting individual employees via their public LinkedIn pages.
Two categories of participants were identified for this study: managers and workers. This
distinction allows separate examination of the effects of the 4DW on the production and
consumption sides of the economy, in accordance with the framework explicated in Section
3.
Managers are defined as senior figures that played a role in overseeing their workplace’s
move to the 4DW, and are able to speak to the organisation’s strategy, management, and
economic performance throughout the transition.
Workers are defined as people who are employed at organisations that have implemented the
4DW model, but did not play a management role through the transition.
Participants were categorised as either managers or workers based on their position title. In
cases where this was not immediately obvious, participants were invited to self-identify as
either a manager or worker according to the above descriptions.
Selection criteria were used to ensure the robustness of the sample. First, at the time of
interviewing, all participants were working at an organisation in the UK that had transitioned
from a five-day full-time schedule to a 4DW without loss of pay. Second, participants were
required to have predominantly resided in the UK for the duration of the 4DW transition to
ensure comparability between consumption data.
Ten managers were interviewed – one from each of ten different organisations. Among them
were five CEOs, two directors, a managing director, a financial director, and a creative
director. The organisations at which these managers work show striking sectoral diversity,
and include two from the non-business sector, three that could be described as sector-specific
40
consultancies11, and one from each of advertising, manufacturing, education, care, and
environmental consulting. The sizes of the organisations ranged from 7 to roughly 650
employees, with the median lying at 18.
Fourteen workers were interviewed across nine different organisations.
Section 4.2.2: Interview process
Interviews were conducted in a semi-standardised format: a series of approximately 12
predetermined questions were asked to each participant in a “systematic and consistent order”
(Lune and Berg, 2017, p. 69). Whilst the interview process was designed to obtain data on
specific phenomena relevant to the analytical framework, participants were encouraged to
provide as much detail as they saw fit and digress to cover other conceptual ground that they
deemed relevant.
Separate sets of questions were developed for the manager and worker samples. Interviews
with managers were designed to explore the impacts of the 4DW on production. As such,
these interviews focussed on impacts related to productivity, revenue, output, workplace
management, energy usage, business transport, staffing, technology, and organisational
carbon footprint.
Interviews with workers were designed to mostly examine the effects of moving to a 4DW on
consumption. Accordingly, workers were asked about the impacts on their time use and
spending patterns, travel patterns and personal carbon footprints. Given that income has been
shown by previous studies to be a key determinant of household emissions (Fremstad et al.,
2019; Nässén and Larsson, 2015), workers were asked about their willingness to trade their
income for leisure time. Furthermore, interviews also explored the impact of the 4DW on
productivity from the workers’ personal experiences, enabling triangulation of this variable
by drawing on both managers’ and workers’ perspectives (Carter et al., 2014).
All interviews took place online and lasted 25-45 minutes. Each participant gave their written
and verbal consent to be interviewed and have audio recorded12.
11 More specific descriptions of these companies are omitted to protect their identities.
12 CUREC approval was granted, with reference number: SOGE C1B 23 21.
41
Section 4.2.3: Data analysis
All interviews were audio recorded. These recordings were transcribed using the transcription
feature in Microsoft Word Online. Data were then deductively categorised and analysed
based on the analytical framework in Section 3 and the conceptual scheme developed in the
WTR literature.
Two of the fourteen workers interviewed were either working part-time or unemployed prior
to beginning the 4DW. Similarly, one manager works at a start-up that implemented the 4DW
from the company’s inception. In these instances, it is impossible to establish baseline values
for the variables of interest to this study, and as such, data from these three interviews are
excluded from parts of my analysis.
Due to the nature of this research methodology, the results obtained through these interviews
often represent participants’ best guesses of the variable in question. For example, given that
comments regarding spending were not cross-checked against personal bank account data,
workers’ answers ultimately measure their subjective impressions of the impact, not the
impact itself. Nevertheless, unless otherwise stated, participants’ responses are taken at face
value and assumed to adequately capture the effect in question.
Section 5: Interview results
In this section, I present results from the research interviews. The findings are arranged in
separate subsections for production and consumption to examine the impacts on each side of
the economy.
Section 5.1: Production
Section 5.1.1: 4DW format
Nine of the ten organisations involved in this study switched to the 4DW in June 2022 as part
of the official 4-Day Week Pilot programme. One organisation was not part of this
programme and adopted the 4DW in January 2023.
42
A reduction in official working hours by roughly 20% was achieved at all ten organisations
by eliminating one workday from the week. In five of the ten workplaces, all employees now
take Fridays off, effectively shutting the business for a three-day weekend. In the other five,
employees are given a choice of their day off, usually between Monday or Friday, with days
off coordinated to ensure that the business can remain open for the whole week.
Several managers noted that the most difficult part of the 4DW transition was coordinating
with business partners and clients who still follow the five-day week schedule.
Section 5.1.2: Productivity and output
Each manager and worker reported significant increases to productivity, understood here as
output per hour, from moving to the 4DW. In many cases, productivity gains were estimated
to be proportional to the reduction in official working hours (Fig. 5), implying that the 4DW
had no net effect on output or revenue. Indeed, firm-level output and revenue either remained
stable or increased in every organisation.
In one third of interviews, managers, who were asked about firm-level productivity, reported
that productivity increased by an amount more than proportional to the reduction in hours
(Fig. 5) – that is, total output or revenue increased as a result of the 4DW. For example, the
manager of the education provider reported productivity growth of 30-35% attributed wholly
to the 4DW, resulting in a 10% increase in sales without changing the number of employees.
The biggest productivity increase occurred in the manufacturing company, where total output
grew by 40% following the move to the 4DW.
Interestingly, workers, who were asked about their personal productivity, were more likely to
report productivity increases that were less than proportional to the official reduction in
working hours (Fig. 5). This was evident in that some workers said that they may work an
hour or two on their day off to finish up all their work for the week.
43
Figure 5: Magnitude of the increase in productivity resulting from the 4DW, split by
responses from managers (n=9) and workers (n=12).
Several major reasons were offered to explain the substantial productivity growth observed.
Two of these were cited by almost every participant: first, productivity was enhanced by
certain psychological and motivational effects specific to the 4DW; and second, the
implementation of the 4DW was invariably accompanied by concomitant changes to work
processes designed to improve efficiency and make the 4DW possible. These two
mechanisms are understood to account for the majority of the observed productivity gains.
Besides these two reasons, some managers explained their productivity gains with reference
to new capital investments or modifications to their product. Discussion of capital
investments is reserved for a different subsection, but I will expand on each of these other
points below.
Psychological effects related to increased happiness and motivation were commonly cited as
a key mechanism behind productivity growth. Both workers and managers frequently
explained productivity gains with reference to what I term “deadline effects” – the increased
motivation that comes when time is scarce due to bringing forward a deadline:
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Managers Workers
Proportion of responses
Less than proportional Proportional Greater than proportional
44
“…it’s being aware that you have fewer days to complete the same
thing…it’s comparable [to] if you have a report or homework piece and the
professor gave you one week less to complete it, you will still complete it in
time, and it could be to the exact same standard, but you have a bit more
pressure on you to do it in time.”
– Worker no. 8
“It’s amazing how quickly things can get done whenever time is limited.”
– Manager no. 2
Many managers noted that employees seemed happier and more engaged following the
transition, which was echoed by workers themselves and cited as another factor behind
increased productivity.
Eight of the nine managers reported that workplace protocols were modified during the 4DW
transition to eliminate inefficiencies and enhance productivity. Cutting the frequency and
length of meetings, reducing messaging and email checking, and rearranging schedules to
group similar tasks together were some of the principal changes implemented to this end.
Participants generally indicated that the workplace changes implemented to facilitate the
4DW would not have increased productivity by the same magnitude were they instituted
under a five-day week. For example, cutting meetings and taking shorter breaks has resulted
in work time becoming more intensive. Employees commented that they are able to deal with
this increased work intensity because they know that they now have an extra day off to
decompress. In this sense, the 4DW makes it possible to work more intensively, and working
intensively makes it possible to implement the 4DW:
“We could have [implemented the management changes in a five-day week
context], but I don’t think there would have been the motivation there to do
it.”
– Manager no. 10
45
Section 5.1.3: Energy and transport
Energy consumption, aggregated from on-site workplace energy use, business transport (i.e.
excluding regular commuting), and commuting, experienced a small but definite reduction
across the nine organisations. Given no change to output (see Section 5.1.2), this implies that
the energy intensity, and hence carbon intensity, of production decreased due to the 4DW.
This supply-side decarbonisation reduces the emissions embodied in goods that are
eventually consumed, putting downwards pressure on the carbon intensity of consumption
and leisure (?% and ?& from Section 3).
Energy use
Six managers reported no change to the organisation’s energy consumption due to the 4DW.
The absence of any effect was generally attributable to these companies’ offices still being
staffed Monday to Friday, or being based remotely or in shared offices.
Two managers estimated decreases in energy consumption due to their offices shutting down
on Fridays. Conversely, energy consumption at the manufacturing business actually increased
slightly, as efficient reorganisation of work allowed them to run the machines for longer and
increase output.
Business transport
The majority of workplaces follow a hybrid or remote working arrangement, and as such
their business transport usage was negligible to begin with. The environmental consultancy
and the healthcare business, however, both track milage for site visits and found significant
decreases resulting from the 4DW. For example, milage at the environmental consultancy
decreased 32% year-on-year due to the 4DW.
Commuting
The effect of the 4DW on workers’ commuting travel was negative, but very small given that
every worker interviewed was already on a hybrid or fully remote working arrangement.
Three hybrid workers reported that they may commute slightly less now that the workweek is
just four days, but two of these three take public transport to work anyway, implying that
46
their commute is already low carbon. This leaves just one worker who may have occasionally
driven to the office on Friday but now does not. The impact on the commuting emissions of
the workers sampled is thus presumed to be virtually zero.
Section 5.1.4: Capital and investment
Two of nine managers reported new capital investments associated with adoption of the
4DW. The manufacturing firm introduced new technology to increase product yield, and the
environmental consultancy invested in tablet technology and new software to speed up on-
site data entry.
Both managers indicated that these investments would have likely occurred at a later date
regardless of the 4DW, but that the productivity drive associated with the transition gave
them cause to bring forward these investments.
Section 5.1.5: Staffing
Each manager reported that no new staff were hired as a direct result of adopting the 4DW,
nor were any employees laid off.
Section 5.1.6: Carbon footprint
Five of nine managers estimated that their organisation’s carbon footprint would have
decreased as a result of the 4DW. The rationale for these answers generally lay in the
perceived reductions in commuting and/or office energy consumption.
The other four managers estimated no change in this variable.
Section 5.2: Consumption
Section 5.2.1: 4DW format
Twelve of the fourteen workers interviewed have their scheduled day off on Friday. One
worker has their day off on Monday, and another one has the freedom to choose their day off
on a week-to-week basis.
47
Section 5.2.2: Consumption and spending
Scale
The scale of workers’ aggregate weekly spending did not exhibit any significant changes due
to the 4DW (Fig. 6), which aligns with the assumptions made regarding income and spending
in Section 3. Seven of twelve workers reported no change to their aggregate spending, and
the changes that were reported in the other five workers were often noted to be small, and
showed no clear directional trend (Fig. 6).
The two workers who were not on a five-day week prior to the official adoption of the 4DW
noted that if required to work a five-day week schedule, their aggregate spending would
likely increase.
Figure 6: Changes to the scale of workers’ aggregate weekly spending following adoption of
the 4DW (n=12). The size of each sector represents the proportion of workers that gave each
corresponding answer.
Composition
Due to compositional changes in spending patterns, demand-side effects of the 4DW on the
carbon intensity of consumption (?%) are most likely positive, creating upwards pressure on
Decreased spending No change Increased spending
48
emissions. Note that this does not account for supply-side effects, which I discuss further in
Section 6.1.
The reasoning for this conclusion is that although aggregate consumption remained roughly
constant, increases to spending often came in the form of consuming additional goods and
services (Table 3a), whereas decreases to spending were more likely to be achieved by
finding lower prices for goods and services that would have been consumed anyway, rather
than people choosing to go without certain products (Table 3b).
For example, many workers responded that they spend more on going out, travel, and
pursuing hobbies under the 4DW. These purchases are additional, in that they would not have
occurred in a BAU scenario. On the other hand, the item that workers most frequently
reported spending less on was food shopping. This reduction in spending, however, did not
occur because workers bought less food, per se, but because workers used their extra day off
to seek out cheaper prices and buy food in bulk.
A similar phenomenon was reported by one worker whose spending on weekend train travel
declined because they began taking trains during off-peak times on a Friday, not because of
any actual decrease to the number of journeys they took.
Thus, the extra time off allows workers to purchase certain items at lower prices, the savings
from which are then used to consume additional goods and services. Workers are thus able to
purchase a higher quantity of goods and services for the same aggregate price, increasing the
material and carbon intensity of each pound spent.
49
Tables 3a (left) and 3b (right): Activities that workers reported spending more money on
(Table 3a) and less money on (Table 3b) following adoption of the 4DW. Frequency refers to
the number of workers (out of twelve) that reported changes to spending. Activity categories
were constructed inductively based on workers’ responses to open-ended questions regarding
their spending patterns.
Increased spending
Decreased spending
Activity
Frequency
Activity
Frequency
Going out, socialising,
eating out
5
Food shopping
3
Leisure travel
3
Going out, socialising,
eating out
3
Exercise
2
Commuting
1
Clothes shopping
1
Leisure travel
1
Hobbies
1
Education
1
Gigs
1
Home improvements
1
Food shopping
1
Dog care
1
In terms of emissions, reduced spending on commuting was reported by just one worker
(Table 3b), and is probably the only significant channel that would depress carbon intensity
of consumption, given that going out is presumed to be relatively low carbon, and that
reduced spending on leisure travel and food shopping arose via cheaper prices, not lower
demand.
On the other side of the equation, the set of items that workers reported spending more on
exhibit heterogeneity in terms of carbon intensities (Table 3a). Items like gigs and gym
memberships are probably low in carbon intensity, whereas leisure travel and clothes
shopping are likely to be much more carbon-intense.
50
Section 5.2.3: Leisure travel
Six of twelve workers reported increases their leisure travel resulting from the 4DW, while
the other six reported no change. Leisure travel emissions likely increased due to the 4DW,
resulting from an uptick in air travel and driving, but this effect may have been slightly
attenuated by a shift in other workers from driving to public transport.
Two workers reported that they take more flights for leisure travel under the 4DW than they
otherwise would. Due to its high carbon intensity, even a small increase in air travel may
cause significant growth in emissions (DESNZ, 2023).
Three workers commented that they drive more often for leisure travel under the 4DW than
previously, while one reported the opposite. The latter case presents an interesting
phenomenon – this worker reported an increase to domestic intercity travel under the 4DW,
but that where this travel may have previously relied on driving, they now prefer to use
public transport because of their newfound time affluence.
Some workers reported that all their leisure travel, both before and after the introduction of
the 4DW, occurs via non-aviation public transport. The additional emissions produced by
increases in leisure travel reported by this cohort are therefore presumed to be very small
(DESNZ, 2023).
Section 5.2.4: Time use changes
It is likely that the carbon intensity of leisure time (?& from Section 3) declined due to the
4DW, driven by a reduction in the consumption intensity of leisure time.
The changes to workers’ leisure time use patterns following the adoption of the 4DW
exhibited remarkable heterogeneity. Resting, socialising, exercising, and completing chores
were among the most frequently-cited activities that workers spend more time on following
the move to the 4DW (Fig. 7).
51
Figure 7: Impact on worker participants’ time use patterns following the adoption of the
4DW (n=12). The y-axis represents the proportion of workers who indicated that they spend
more time doing certain activities compared to before the 4DW. The time use categories on
the x-axis were constructed inductively from workers’ responses to open-ended questions
regarding changes to their time use patterns.
Many workers seem to treat their new day off (usually Friday) as a day to relax and organise
their personal affairs before the traditional Saturday-Sunday weekend starts. Indeed, many
appear to conceptualise of their Fridays as a sort of rest day before the weekend, rather than
an extra weekend day.
Some workers also noted that they now enjoy a slower, more deliberate approach to their
leisure time in general, when previously they may have gone out partying and/or tried to
cram as many activities into their two-day weekend as possible. This shift seems to arise
from psychological effects, in that the extra day of rest can remove the need to “blow off
steam” (Worker no. 12) and encourages a more systematic approach to managing one’s time
budget.
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
Chores
Exercise
Rest
Socialising
Time in nature
Trips (non-flying)
Family time
Volunteering
Hobby
Trips (flying)
Education
Proportion of workers
52
The carbon intensity of leisure time likely declined as a result of these shifting time use
patterns. Recall that the carbon intensity of leisure time is calculated as the product of the
carbon intensity of leisure consumption and the consumption intensity of leisure time
(Section 3.2.2, Eq. 8):
?&(>&
4&A4&
5
The carbon intensity of leisure consumption (0#
%#
) probably remained stable or increased.
While some shift towards low carbon leisure was observed (e.g. switching to public
transport), this was likely outweighed by increased purchasing of carbon-intensive leisure
goods (e.g. flights, petrol, clothes), particularly given the high carbon intensity of air travel
(DESNZ, 2023).
Conversely, the consumption intensity of leisure time (%#
&) likely experienced a moderate
decrease due to the 4DW. This can be explained by the fact that workers generally indicated
that their extra day off tends to be quieter and slower (i.e. less consumption-intensive) than
their conventional two-day weekend, which implies that even though total leisure
consumption (4&) may increase, it probably increases by less, proportionally, than the
increase in leisure time (5).
Overall the decrease in consumption intensity of leisure time (%#
&) likely outweighs any
growth in the carbon intensity of leisure consumption (0#
%#
). Hour for hour, workers’ leisure
time is therefore estimated to be less carbon-intensive under the 4DW than previously,
implying a decrease in ?&.
Section 5.2.5: Trade-off between income and leisure time
All workers bar one expressed that if necessary, they would accept a pay cut to be able to
work a four-day week schedule, however the size of this theoretical pay cut varied between
interviewees. This result indicates that society’s optimal balance between consumption and
leisure, represented by : in Section 3, is not reflected in the prevailing five-day week
paradigm, as discussed further in Section 6.
53
Workers were asked to nominate the fraction of their income that they would be willing to
sacrifice, if necessary, to move to a four-day week schedule. It should be noted that a pay cut
of anything less than 20% still implies an increase to the hourly wage rate, assuming hours
are reduced by exactly 20%.
The most common answer was 10%, given by four of twelve workers. This was also the
median answer. Several other workers also quoted figures around this mark, for example
“between 10-20%” or “5-10%”.
If answers of the form “5-10%” are assumed to take the middle value of the given range (i.e.
7.5% in this example), then the mean salary reduction that workers would be willing to
accept is 11.67%.
Two of twelve workers indicated that they would be willing to accept a 20% pay cut in
exchange for a four-day week, effectively forgoing a full day’s worth of income.
One worker was even willing to accept a 25% salary reduction, implying a decreased hourly
wage, to be able to work a four-day schedule. This strong preference for leisure time over
income was echoed by one of the workers that was not previously working a five-day week
before the official 4DW began; they indicated that they would reject a 25% salary increase if
it were tied to moving to a five-day week.
Only one worker expressed that they were wholly unwilling to trade off their income for
leisure time.
On the whole, these results emphatically show that workers would be willing to sacrifice
some portion of their income in exchange for extra leisure time.
The reasons that workers quoted for their responses were diverse, but a few common themes
can be identified.
54
First, many workers, especially those who expressed willingness to take larger pay cuts,
commented that their extra leisure time is extremely valuable, and has delivered better health
and quality of life:
“What I’ve come to realise is that time is the real most valuable thing that
you can have.”
– Worker no. 7
Second, workers frequently remarked that their appreciation for their leisure time has grown
since the advent of the 4DW. Several commented that pre-4DW, they may have been
reluctant to trade off their income for an extra day off, but that they would be far more likely
to do so now, having experienced its benefits.
Third, the reasons offered by workers as to why they would not be willing to accept larger
pay cuts revolved around the unaffordable cost of living, career progression concerns, and the
perception that taking a pay cut would symbolise some sort of demotion.
Section 5.2.6: Day off preferences
Workers almost uniformly prefer their day off to be scheduled for either Friday or Monday,
with Friday generally preferred to Monday.
Two major reasons were cited to explain this preference: first, leisure time is more restful and
enjoyable when taken consecutively; and second, a midweek day off would interrupt the flow
of their work.
Workers also expressed that their appreciation for their leisure time would be enhanced
further still if their friends and family also had the same day off.
Section 5.2.7: Carbon footprint
Seven of the twelve workers felt that their personal carbon footprint did not change as a result
of moving to the 4DW.
55
Three workers felt that it decreased, one of which attributed this to commuting less
frequently, another to their shift to a less consumption-intensive lifestyle, and the third to a
switch from driving to public transport, made possible by their newfound time affluence.
Two workers felt that their carbon footprint grew due to the 4DW – both of which attributed
this to increased travel.
Section 6: Discussion
In this section, I first describe my key findings by deploying the analytical framework
developed in Section 3 to interpret the interview results. I discuss these findings in relation to
the literature, before assessing the long-term impact of the 4DW on emissions based on
productivity trends. I then discuss implications for policymaking, and conclude by outlining
the limitations of this study.
Section 6.1: Key findings – does the 4DW decrease emissions?
Section 6.1.1: Interpretation of interview results using the analytical framework
Based on the interview results and the emissions model (Eq. 7, see below), the 4DW is likely
to produce a small increase in emissions in the short term driven by growth in leisure
emissions. However, this result is subject to significant uncertainty regarding the magnitude
of changes to different variables.
> ( ?%4 @?&5
Consumption emissions as defined in the above equation (?%4) would probably decline
slightly following the adoption of the 4DW due to a slight reduction in the carbon intensity of
consumption (?%). Based on the interview results, aggregate consumption (4) is unlikely to
change as a result of the 4DW with no loss of pay, which aligns with the assumptions made
regarding spending and income in Section 3. Meanwhile, the carbon intensity of consumption
(?%) may decrease slightly; however, this result is less certain. Based on interview results,
shifting consumption habits place upwards pressure on the carbon intensity term ?% from the
demand side (see Sections 5.2.2 and 5.2.3). Meanwhile, the workplace decarbonisation
effects described in Section 5.1.3 reduce the carbon intensity of goods, creating downwards
56
pressure on ?% from the supply side. On balance, the supply-side effects probably dominate
given, for example, the significant decreases reported in business transport milage, implying
a small net decrease in ?% overall. Interestingly, this conclusion is reflected in interviewees’
estimates of their carbon footprints: five of nine managers estimated that their organisation’s
carbon footprint declined due to the 4DW, whereas only three of twelve workers felt this way
about their personal carbon footprint. This estimate is, however, highly uncertain given the
difficulties in accurately measuring changes in consumption habits and the diffuse impacts of
supply-side decarbonisation on consumption emissions. Nevertheless, assuming a slight
decrease to ?% and no change to 4, I estimate that a slight decrease in consumption emissions
(?%4) is likely to occur under the 4DW.
Conversely, leisure emissions (?&5) would likely experience a moderate increase as a result
of the 4DW. Leisure time (5) increases substantially, which is the point of WTR. Based on
the interview results, the carbon intensity of leisure time (?&), however, would likely
decrease. Whilst growth in carbon-intensive travel threatens to increase ?&, this is
outweighed by two effects that act in the opposite direction. First, the consumption intensity
of leisure time is likely to decrease (see Section 5.2.4). Second, supply-side decarbonisation
reduces the carbon intensity of leisure goods that are consumed. However, although people’s
leisure time may be, hour for hour, less carbon-intensive under the 4DW than previously, the
large increase in leisure time afforded by the 4DW means that total leisure emissions would
likely increase.
I estimate that this moderate increase in leisure emissions is greater than the small decrease in
consumption emissions described above, producing a slight short-term increase in total
emissions resulting from the 4DW, according to the model. Key uncertainties remain,
however, regarding the balance of supply-side decarbonisation effects and the various
demand-side effects that together determine the carbon intensity terms.
This estimate is plausible based on findings from other household-level studies. Nässén and
Larsson (2015) arrive at an almost identical conclusion, estimating a very slight increase in
household consumption emissions associated with WTR, net of income. Conversely, my
result differs from that of Fremstad et al. (2019), who estimate composition effects to be
small and positive based on household-level data. However, given the large methodological
57
differences and the uncertainty of my estimate, due in large part to the small sample size and
qualitative interview approach, it is difficult to draw meaningful comparisons with previous
studies.
Ultimately, my findings suggest the need for caution in forecasting potential emissions
reductions from WTR, echoing similar conclusions from Hanbury et al. (2019) and Shao and
Rodríguez-Labajos (2016).
58
Table 4: Empirically-determined effects of the 4DW on different variables.
Variable
Effect of 4DW based on
interview data
Notes
Labour &
Decrease
By definition
Leisure time 5
Increase
By definition
Consumption 4
No change
No reported change
Ratio of consumption
to leisure time
%
&
Decrease
Stable consumption and increased
leisure time implies a decrease in this
ratio
Productivity )
Increase
Drastic increase
Wage rate /
Increase
Increase is proportional to decrease in
working hours; made possible by
productivity growth
Capital '
Increase
Small increase
Output/GDP %
No change
No reported change
Carbon intensity of
consumption ?%
Decrease
Slight decrease driven by supply-side
decarbonisation; subject to high
degree of uncertainty
Carbon intensity of
leisure ?&
Decrease
Large increase in leisure time implies
reduction in per hour carbon intensity
Consumption
emissions ?%4
Decrease
Slight decrease due to reduction in
carbon intensity of consumption
Leisure emissions
?&5
Increase
Moderate increase driven by large
increase in leisure time
Emissions >
Increase
Slight increase driven by leisure
emissions growth, subject to high
degree of uncertainty
Before concluding this subsection, it is necessary to briefly explain the relationship between
my analytical framework and the established approach in the literature. The conventional
approach assumes consumption to be the sole source of emissions, and assesses changes to its
scale and composition to estimate impacts. In this dissertation, I employ a different approach
by going beyond consumption to also incorporate supply-side effects. I separate emissions
59
into those from consumption and those from leisure, which facilitates the linkage between
production and consumption via the equilibrium modelling in Section 3. In this formulation,
the scale effect is visible only in consumption emissions (via 4), whilst separate composition
effects are defined for each of consumption (?%) and time use (?&). There is no direct scale
effect (i.e. via income) on leisure emissions, but leisure emissions do vary with leisure
consumption via ?& (see Section 3.2.3, Eq. 8). Thus, my lens of analysis is distinct from, but
related to, the conventional approach described in the literature.
Section 6.1.2: Looking beyond the analytical framework
This estimate of a small, short-term increase in emissions is validated by analysis using the
conventional lens of scale and composition effects. The 4DW was confirmed to affect
emissions via composition effects alone given workers’ incomes and aggregate spending
remained unchanged, implying zero short-term scale effects. Based on interview data,
composition effects are difficult to determine with any degree of certainty given the various
effects acting in opposition to one another, as described above, but are likely to be slightly
negative, implying a small increase in emissions.
This result differs from those of the macrostructural analyses that estimate composition
effects to be positive and statistically significant (Fitzgerald et al., 2018; Hayden and
Shandra, 2009; Mallinson and Cheng, 2021), which can be explained by two reasons. First,
these papers use different metrics – either ecological footprint or territorial carbon emissions,
not consumption emissions, which was the frame of analysis for this dissertation. Indeed, the
only macrostructural study to look at carbon footprint found no statistically significant
composition effect (Knight et al., 2013). Second, this discrepancy may reflect the hypothesis
that countries are more effective at reducing emissions via WTR than households due to
social multiplier effects, resulting in larger estimates of composition effects in macro-level
studies compared to those at the household level.
This dissertation sheds new light on the effects of WTR on emissions via supply-side
decarbonisation, finding a clear reduction in the carbon intensity of production via decreased
business transport and on-site energy use. Interestingly, the reduction of commuting
emissions due to the 4DW was negligible in the workers sampled, but the potential size of
this effect is likely underestimated. Most UK workers do not work from home at all (ONS,
60
2023b), and given that 68% of commuting trips occur by car (DfT, 2022), decreased
commuting from the 4DW would likely be a significant source of supply-side emissions
reductions that was not captured. The absence of this effect is probably explained by an
overlap in the kinds of workplaces that follow remote or hybrid working arrangements and
those that have implemented the 4DW, or are able to.
The modelling and analysis in this dissertation has, to this point, centred on the short-term
impacts of the 4DW on emissions. It is possible, however, that scale effects may arise from
the 4DW over time and reduce emissions compared to BAU via avoided economic growth,
provided future productivity and wage growth under the 4DW is slower than that under BAU.
Under BAU, working hours are held roughly stable while productivity gains are converted
into increased incomes and economic growth. Productivity-led WTR, conversely, translates
productivity growth into increased leisure time rather than higher incomes, thus reducing
output and emissions compared to BAU via avoided economic growth (Cieplinski et al.,
2021b). In instances of gradual productivity-led WTR, this effect is relatively
straightforward: labour becomes more productive which facilitates a reduction in working
hours with no loss of pay.
In the case of the 4DW however, organisations embark on a productivity blitz to make the
transition possible. Much of this productivity growth is additional, in that it would not have
occurred under BAU (see Section 5.1.2). Herein lies the key point: if these productivity gains
would have never occurred under BAU, then there are no avoided emissions associated with
the 4DW, per se. To use a concrete example, if meetings and emails were never going to be
reduced under BAU, then the productivity gains from these efficiency measures were never
going to result in increased incomes and economic growth. If, alternatively, productivity
gains result from BAU measures that are brought forward to facilitate the 4DW transition, as
in the cases of the capital investments made by the environmental consultancy and the
manufacturing company, then the conversion of this productivity growth into extra time off
would represent some amount of avoided emissions. Where this occurs, the 4DW effectively
represents firms leveraging future productivity growth to smooth the transition to a less
carbon-intensive society, all the while fostering social acceptability from the perspectives of
both workers (via stable incomes) and firms (via stable output and profits). Under this
scenario, even if emissions remain stable or increase in the short term (e.g. due to
61
composition effects), a decrease compared to BAU may arise in the long term via avoided
economic growth.
The 4DW thus only reduces emissions compared to BAU via the scale effect insofar as it
translates BAU productivity gains into leisure time. The extent to which this actually occurs
is ultimately very difficult to measure, and probably best done so by comparing future
productivity growth in 4DW firms to other comparable organisations in the same industry.
Based on my interview data, most of the productivity growth achieved during the 4DW
transition is ostensibly additional (i.e. non-BAU) – in that it stems from either effects unique
to the 4DW (e.g. well-rested workers) or from management measures specifically
implemented to facilitate the transition (e.g. cutting down emails). The example of capital
investments described above perhaps represents the only instance of future productivity gains
being brought forward, and hence may imply a small amount of avoided economic growth
and emissions.
Ostensibly in contrast to this result, Cieplinski et al. (2021b) model significant emissions
reductions over time resulting from productivity-led WTR. These divergent conclusions arise
from differing explanations for productivity growth. Cieplinski et al. (2021b) assume
productivity gains occur only via technological improvements, and are then translated into
extra leisure time. In contrast, I empirically determine the sources of productivity growth
under the 4DW to be, for the most part, endogenous to WTR, implying little in the way of
avoided economic growth and emissions compared to BAU. Interestingly, the small portion
of observed productivity growth that I argue reduces emissions compared to BAU arises from
technology upgrades, thus supporting the findings of Cieplinski et al. (2021b) regarding the
relationship between technology, productivity, WTR, and emissions.
Thus, my findings regarding productivity represent an original contribution to the literature
and a principal component of this dissertation. First, productivity was shown via theoretical
analysis to be a key determinant of wages, output, and emissions under WTR (see Section
3.4). Second, empirical investigation revealed significant productivity gains associated with
the 4DW transition, and, crucially, their sources (see Section 5.1.2). Lastly, in the above
discussion, I have analysed the nature of the observed productivity gains and drawn novel
conclusions regarding future trends in productivity, emissions and economic growth. It is
only via the integration of modelling and qualitative interview methodologies employed in
62
this dissertation that I am able to draw out these insights. Indeed, to the best of my
knowledge, I am the first to explicate the relationship between productivity, WTR, and
emissions in this fashion.
Section 6.2: Policy discussion
In this section, I discuss the implications of my findings for WTR policymaking.
First, the 4DW cannot be assumed to automatically reduce emissions, and should not be
treated as such. Given no change to the scale of consumption, the emissions impacts of the
4DW are determined by changes in the composition of consumption. If WTR is to reduce
emissions, policy should thus aim to promote low carbon leisure and attenuate carbon-
intensive consumption. A carbon tax or similar levy could shift consumption away from
carbon-intensive activities, such as flying, and has been shown to have significant
complementarity with WTR (Cieplinski et al., 2021a). It’s also possible that deliberate
messaging around WTR programmes that emphasises rest over consumption, or an
environmental framing, may promote this shift.
Second, and relatedly, WTR policy could be crafted to favour low carbon leisure activities.
For example, government policy could support workers to reduce their hours on the condition
that their time off is spent on some low carbon activity, such as parenting, further education,
or volunteering. This kind of policy seems plausible: similar programmes already allow
young Swiss employees to take leave from work for youth activities and sports (SECO,
2020).
Third, based on my interview results, a large proportion of workers are willing to forgo some
portion of their income for extra leisure time, and should be supported to do so, especially in
workplaces where the 4DW without loss of pay is not feasible. This is a particularly
important point, as it indicates that our prevailing sociolegal institutions do not represent the
socially optimal balance between consumption and leisure. In other words, society’s true
value of :, which describes the preference for consumption over leisure in the household
utility function from Section 3, is lower than that reflected in the current five-day week
arrangement. Workers essentially want to live less consumption-intensive lifestyles, but are
63
prevented from doing so by current policy arrangements. This finding has profound
implications for WTR policymaking in the UK.
Policy should make it as easy as possible for workers to request part-time schedules (i.e.
WTR with a proportional pay cut), and promote the conversion of productivity gains into
leisure time, rather than higher incomes, wherever possible. Facilitating this reorientation
requires dismantling the norms and institutions that have historically encouraged employers
to offer higher wages in lieu of shorter working hours, such as those described by Schor
(2005). For example, labour market regulations could allow workers to request fewer hours
instead of a pay raise, as in Austria (Gerold and Nocker, 2018). Essentially, policy should
aim to facilitate the trade-off between income and leisure in situations where workers are
overemployed. The social impact of this can be mitigated by agreements involving wage
freezes, rather than cuts, as people are more willing to forgo future pay raises than accept a
cut today (De Spiegelaere and Piasna, 2017). A programme of this nature offers tremendous
upside in both enhanced wellbeing and emissions reductions via scale effects.
One challenge to productivity-led WTR in the UK context, however, is the UK’s secular
productivity slowdown: productivity growth in the UK averaged just 0.6% per year over the
decade 2010-2020 (ONS, 2022). At this rate, under a gradual productivity-led WTR
programme, it would take 38 years to reach a four-day week with no change to incomes,
assuming zero endogenous productivity growth under WTR. Incremental reductions in hours,
however, may actually represent a smoother low carbon transition compared to the abrupt
introduction of the 4DW studied in this dissertation. Moreover, given that many workers are
willing to sacrifice some portion of their income for shorter working hours, the rate of WTR
is not necessarily constrained by the wage growth rate. In any case, there are key trade-offs
between productivity, emissions, incomes, and economic growth that must be co-managed by
workers, employers, and policymakers.
Fourth, where possible, policy should aim to coordinate WTR at a societal level and promote
a common day off, ideally Friday. From an environmental perspective, arranging a society-
wide day off on Friday is likely to bring larger emissions reductions compared to other WTR
arrangements due to coordinated reductions in commuting and workplace energy use, as
reflected in my interview results and King and van den Bergh’s (2017) theoretical scenario
analysis. From a wellbeing perspective, workers interviewed for this dissertation almost
64
unanimously expressed a strong preference for their day off to be Friday, and some also
noted that their appreciation for their day off would be enhanced further still if their friends
and family also had Fridays off. Lastly, from an employer perspective, several managers
noted that the most difficult aspect of the 4DW transition was coordinating between other
business partners and clients who are not on a four-day schedule, an issue which would
diminish if the day off was coordinated across society.
Section 6.3: Limitations and future directions
This study would have benefitted immensely from access to detailed spending and time use
data for workers on the 4DW. Given that this data was unavailable for workers sampled in
this study, my conclusions instead rely on qualitative interview data, resulting in a high
degree of uncertainty surrounding my estimates of carbon intensities, as described in Section
6.1. The imprecision of these estimates is a key limitation of this dissertation. Access to
detailed personal data, coupled with a more systematic approach to carbon footprint
estimation, would enhance the accuracy and precision of estimates in future research.
Moreover, this study is the first to empirically investigate the supply-side effects of WTR on
emissions, yielding several key insights in this regard. These effects are understudied, and
future research should focus on determining their size, mechanisms, impact on overall carbon
intensity, and implications for policy and business management.
Furthermore, this dissertation’s estimates of supply-side decarbonisation effects come with
the caveat that the impact of capital investment on emissions was not modelled. It’s possible
that firms may respond to WTR by investing in machinery to boost productivity, which could
affect emissions. While this effect is probably small in the sampled organisations, the
relationship itself is complex and warrants further investigation.
In this dissertation, I have assumed that slower economic growth ipso facto reduces
emissions, but certain exceptions may exist. For example, if WTR in the renewable energy
industry slowed its growth, emissions may not decline as fast as they would under BAU. The
effects of WTR on emissions are thus likely to exhibit heterogeneity across sectors, offering a
compelling topic for future research.
65
Lastly, this dissertation has not considered the interactions between WTR and income
distribution. Roughly one in eight UK workers currently earns less than the Living Wage
(Aziz and Richardson, 2022), and it is likely that many in this cohort would prefer higher
incomes, either via working longer hours or wage increases, rather than extra time off, but
this phenomenon and its implications were not examined in my analyses. Moreover, the
relationship between working hours and emissions is likely shaped by economic inequality,
implying that WTR and policies that reduce inequality may exhibit synergy in reducing
emissions (Fitzgerald, 2022). Similarly, the impact of WTR on one’s personal carbon
footprint likely varies with income, and future research should aim to investigate this
relationship.
Section 7: Conclusion
Contrary to mainstream discourse, I have shown that adoption of the 4DW with no loss of
pay is likely to increase emissions slightly in the short term. Notwithstanding clear supply-
side emissions reductions, this result is driven by shifting consumption habits – namely,
growth in carbon-intensive leisure consumption and the use of price discounts to expand
consumption baskets. In the long term, however, it is possible that the 4DW could lead to
small emissions reductions compared to BAU via avoided economic growth, but these effects
are difficult to forecast and depend on a slowdown in future productivity and wage growth.
The potential for the 4DW with no loss of pay to reduce carbon emissions in the UK is
therefore limited. However, other forms of WTR are likely to prove more successful in this
regard. I have shown that the current balance of consumption and leisure is misaligned
towards the former, which suggests that significant benefits to both wellbeing and the
environment are possible via reshaping of the sociolegal institutions that govern working
hours. This offers a useful starting point for realising multiple potential benefits of WTR.
Ultimately, as demonstrated in this dissertation, the relationship between working hours and
emissions is complex and shaped by a variety of social and economic factors. If the full
potential of WTR is to be realised, the transition must be attentive to these nuances and
carefully planned to neutralise potential rebound effects.
66
References
Alesina, A., Glaeser, E., Sacerdote, B., 2005. Work and Leisure in the United States and
Europe: Why So Different? https://doi.org/10.1086/ma.20.3585411 20, 1–64.
https://doi.org/10.1086/MA.20.3585411
Angrave, D., Charlwood, A., 2015. What is the relationship between long working hours,
over-employment, under-employment and the subjective well-being of workers?
Longitudinal evidence from the UK. Human Relations 68, 1491–1515.
https://doi.org/10.1177/0018726714559752/ASSET/IMAGES/LARGE/10.1177_001872
6714559752-FIG2.JPEG
Antal, M., Plank, B., Mokos, J., Wiedenhofer, D., 2020. Is working less really good for the
environment? A systematic review of the empirical evidence for resource use,
greenhouse gas emissions and the ecological footprint. Environmental Research Letters
16, 013002. https://doi.org/10.1088/1748-9326/ABCEEC
Aziz, S.A., Richardson, J., 2022. Employee Jobs Paid Below the Living Wage: 2022.
Booth, A., Schiantarelli, F., 1987. The Employment Effects of a Shorter Working Week.
Economica 54, 248. https://doi.org/10.2307/2554393
Bowen, A., Hepburn, C., 2014. Green growth: an assessment. Oxf Rev Econ Policy 30, 407–
422. https://doi.org/10.1093/OXREP/GRU029
Carter, N., Bryant-Lukosius, D., Dicenso, A., Blythe, J., Neville, A.J., 2014. The use of
triangulation in qualitative research. Oncol Nurs Forum 41, 545–547.
https://doi.org/10.1188/14.ONF.545-547
Cieplinski, A., D’Alessandro, S., Distefano, T., Guarnieri, P., 2021a. Coupling environmental
transition and social prosperity: A scenario-analysis of the Italian case. Structural
Change and Economic Dynamics 57, 265–278.
https://doi.org/10.1016/J.STRUECO.2021.03.007
Cieplinski, A., D’Alessandro, S., Guarnieri, P., 2021b. Environmental impacts of
productivity-led working time reduction. Ecological Economics 179, 106822.
https://doi.org/10.1016/J.ECOLECON.2020.106822
Collewet, M., Sauermann, J., 2017. Working hours and productivity. Labour Econ 47, 96–
106. https://doi.org/10.1016/J.LABECO.2017.03.006
Cornwall, A., Jewkes, R., 1995. What is participatory research? Soc Sci Med 41, 1667–1676.
https://doi.org/10.1016/0277-9536(95)00127-S
67
Costanza, R., Alperovitz, G., Daly, H., Farley, J., Franco, C., Jackson, T., Kubiszewski, I.,
Schor, J., Victor, P., 2013. Building a Sustainable and Desirable Economy-in-Society-
in-Nature, Building a Sustainable and Desirable Economy-in-Society-in-Nature : Report
to the United Nations for the 2012 Rio+20 Conference. ANU Press.
https://doi.org/10.26530/OAPEN_462196
Crafts, N., Mills, T.C., 2020. Is the UK productivity slowdown unprecedented? National
Institute Economic Review 251, R47–R53. https://doi.org/10.1017/NIE.2020.6
Cutanda, A., Sanchis-Llopis, J.A., 2019. Intertemporal Substitution for Consumption and
Leisure: Empirical Evidence for Spain (No. WPAE-2019-1909), Working Papers in
Applied Economics.
Daly, H., 2015. Economics for a Full World.
de Fraja, G., 1996. Entrepreneur or Manager: Who Runs the Firm? J Ind Econ 44, 89–98.
https://doi.org/10.2307/2950563
De Spiegelaere, S., Piasna, A., 2017. The why and how of working time reduction.
DEFRA, 2023. UK full dataset 1990 - 2020, including conversion factors by SIC code. UK
and England’s carbon footprint to 2020.
DESNZ, 2023. Conversion factors 2023: condensed set. Greenhouse gas reporting:
conversion factors 2023.
Devetter, F.X., Rousseau, S., 2011. Working Hours and Sustainable Development.
http://dx.doi.org/10.1080/00346764.2011.563507 69, 333–355.
https://doi.org/10.1080/00346764.2011.563507
DfT, 2022. Transport Statistics Great Britain: 2022 Domestic Travel [WWW Document].
Transport Statistics Great Britain. URL
https://www.gov.uk/government/statistics/transport-statistics-great-britain-
2022/transport-statistics-great-britain-2022-domestic-travel (accessed 9.3.23).
Fitzgerald, J.B., 2022. Working time, inequality and carbon emissions in the United States: A
multi-dividend approach to climate change mitigation. Energy Res Soc Sci 84, 102385.
https://doi.org/10.1016/J.ERSS.2021.102385
Fitzgerald, J.B., Schor, J., Jorgenson, A.K., 2018. Working Hours and Carbon Dioxide
Emissions in the United States, 2007–2013. Social Forces 96, 1851–1874.
https://doi.org/10.1093/SF/SOY014
Fremstad, A., Paul, M., Underwood, A., 2019. Work Hours and CO2 Emissions: Evidence
from U.S. Households. https://doi-org.ezproxy-
68
prd.bodleian.ox.ac.uk/10.1080/09538259.2019.1592950 31, 42–59.
https://doi.org/10.1080/09538259.2019.1592950
Gerold, S., Nocker, M., 2018. More Leisure or Higher Pay? A Mixed-methods Study on
Reducing Working Time in Austria. Ecological Economics 143, 27–36.
https://doi.org/10.1016/J.ECOLECON.2017.06.016
Godwin, R., 2023. Work less, live more: is it time to end the five-day week? The Guardian.
Golden, L., 2012. The effects of working time on productivity and firm performance: a
research synthesis paper.
Gorz, A., 1994. Capitalism, Socialism, Ecology. Verso.
Hallegatte, S., Heal, G., Fay, M., Treguer, D., 2012. From Growth to Green Growth - A
Framework (No. 17841).
Hanbury, H., Bader, C., Moser, S., 2019. Reducing Working Hours as a Means to Foster
Low(er)-Carbon Lifestyles? An Exploratory Study on Swiss Employees. Sustainability
11. https://doi.org/10.3390/SU11072024
Hanna, A.S., Taylor, C.S., Sullivan, K.T., 2005. Impact of Extended Overtime on
Construction Labor Productivity. J Constr Eng Manag 131, 734–739.
https://doi.org/10.1061/(ASCE)0733-9364(2005)131:6(734)
Hausman, D.M., 1992. The Inexact and Separate Science of Economics. Cambridge
University Press. https://doi.org/10.1017/CBO9780511752032
Hayden, A., 1999. Sharing the Work, Sparing the Planet: Work Time, Consumption, &
Ecology. Zed Books.
Hayden, A., Shandra, J.M., 2009. Hours of work and the ecological footprint of nations: an
exploratory analysis. http://dx.doi.org/10.1080/13549830902904185 14, 575–600.
https://doi.org/10.1080/13549830902904185
Hickel, J., 2020a. Quantifying national responsibility for climate breakdown: an equality-
based attribution approach for carbon dioxide emissions in excess of the planetary
boundary. Lancet Planet Health 4, e399–e404. https://doi.org/10.1016/S2542-
5196(20)30196-0
Hickel, J., 2020b. Less is More: How Degrowth Will Save the World. William Heinemann.
Huberman, M., Minns, C., 2007. The times they are not changin’: Days and hours of work in
Old and New Worlds, 1870–2000. Explor Econ Hist 44, 538–567.
https://doi.org/10.1016/J.EEH.2007.03.002
IPCC, 2022. Summary for Policymakers. In: Climate Change 2022: Mitigation of Climate
Change. Contribution of Working Group III to the Sixth Assessment Report of the
69
Intergovernmental Panel on Climate Change.
https://doi.org/10.1017/9781009157926.001
Jackson, T., 2016. Prosperity without Growth: Foundations for the Economy of Tomorrow,
2nd ed. Routledge.
Jalas, M., 2002. A time use perspective on the materials intensity of consumption. Ecological
Economics 41, 109–123. https://doi.org/10.1016/S0921-8009(02)00018-6
Kallis, G., Kerschner, C., Martinez-Alier, J., 2012. The economics of degrowth. Ecological
Economics 84, 172–180. https://doi.org/10.1016/J.ECOLECON.2012.08.017
Kapteyn, A., Kalwij, A., Zaidi, A., 2004. The myth of worksharing. Labour Econ 11, 293–
313. https://doi.org/10.1016/J.LABECO.2003.08.001
Keen, S., 2001. Debunking Economics. Pluto Press Australia.
Keynes, J.M., 1930. Economic Possibilities for our Grandchildren.
King, L.C., van den Bergh, J.C.J.M., 2017. Worktime Reduction as a Solution to Climate
Change: Five Scenarios Compared for the UK. Ecological Economics 132, 124–134.
https://doi.org/10.1016/J.ECOLECON.2016.10.011
Kirman, A.P., 1992. Whom or What Does the Representative Individual Represent? Journal
of Economic Perspectives 6, 117–136. https://doi.org/10.1257/JEP.6.2.117
Knight, K.W., Rosa, E.A., Schor, J., 2013. Could working less reduce pressures on the
environment? A cross-national panel analysis of OECD countries, 1970–2007. Global
Environmental Change 23, 691–700.
https://doi.org/10.1016/J.GLOENVCHA.2013.02.017
Lewis, K., Stronge, W., Kellam, J., Kikuchi, L., Schor, J., Fan, W., Kelly, O., Gu, G., Frayne,
D., Burchell, B., Bridson Hubbard, N., Kamarāde, D., Mullens, F., 2023. The results are
in: The UK’s Four-day Week Pilot.
Lune, H., Berg, B.L., 2017. Qualitative Research Methods for the Social Sciences, 9th ed.
Pearson.
Mallinson, D.J., Cheng, K.J.G., 2021. The relationship between state-level carbon emissions
and average working hours in the United States: a replication study. https://doi-
org.ezproxy-prd.bodleian.ox.ac.uk/10.1080/23251042.2021.1975350 8, 88–93.
https://doi.org/10.1080/23251042.2021.1975350
Marimon, R., Zilibotti, F., 2000. Employment and distributional effects of restricting working
time. Eur Econ Rev 44, 1291–1326. https://doi.org/10.1016/S0014-2921(00)00032-5
Mayhew, K., 1991. The assessment: The UK labour market in the 1980s. Oxf Rev Econ
Policy 7, 1–17. https://doi.org/10.1093/OXREP/7.1.1
70
Nässén, J., Larsson, J., 2015. Would shorter working time reduce greenhouse gas emissions?
An analysis of time use and consumption in Swedish households. Environ Plann C Gov
Policy 33, 726–745. https://doi.org/10.1068/C12239
OECD, 2011. Towards green growth: A summary for policy makers.
Oh, S.Y., Park, Y., Bowles, S., 2012. Veblen effects, political representation, and the
reduction in working time over the 20th century. J Econ Behav Organ 83, 218–242.
https://doi.org/10.1016/J.JEBO.2012.05.006
O’Hara, B., 1993. Working Harder Isn’t Working: How We Can Save the Environment, the
Economy, and Our Sanity by Working Less and Enjoying Life More. New Star Books.
ONS, 2023a. EARN01: Average weekly earnings.
ONS, 2023b. Characteristics of homeworkers, Great Britain: September 2022 to January
2023 [WWW Document]. Opinions and Lifestyle Survey. URL
https://www.ons.gov.uk/employmentandlabourmarket/peopleinwork/employmentandem
ployeetypes/articles/characteristicsofhomeworkersgreatbritain/september2022tojanuary2
023#data-sources-and-quality (accessed 9.3.23).
ONS, 2023c. Labour costs and labour income, UK.
ONS, 2022. Labour productivity. Labour productivity: Tables 1 to 8 and R1.
Peters, G.P., Hertwich, E.G., 2008. CO2 embodied in international trade with implications for
global climate policy. Environ Sci Technol 42, 1401–1407.
https://doi.org/10.1021/ES072023K/SUPPL_FILE/ES072023K-FILE002.PDF
Read, S., Hooker, L., Simpson, E., 2023. Firms stick to four-day week after trial ends. BBC.
Roos, C.F., 1935. Economic Theory of the Shorter Work Week. Econometrica 3, 39.
https://doi.org/10.2307/1907344
Rosnick, D., Weisbrot, M., 2006. Are Shorter Work Hours Good for the Environment? A
Comparison of U.S. and European Energy Consumption. SAGE PublicationsSage CA:
Los Angeles, CA. https://doi.org/10.2190/D842-1505-1K86-9882
Schor, J., 2010. Plenitude: The New Economics of True Wealth. Penguin Press.
Schor, J., 2005. Sustainable Consumption and Worktime Reduction. J Ind Ecol 9, 37–50.
https://doi.org/10.1162/1088198054084581
Schor, J., 1995. Can The North Stop Consumption Growth? Escaping The Cycle of Work and
Spend, in: The North, the South, and the Environment: Ecological Constraints and the
Global Economy. Routledge, pp. 68–84. https://doi.org/10.4324/9781315070537-4
Schor, J., 1992. WORK, SPEND, WORK, SPEND: IS THIS ANY WAY TO LIVE?
Washington Post.
71
Schor, J., 1991. Global equity and environmental crisis: An argument for reducing working
hours in the North. World Dev 19, 73–84. https://doi.org/10.1016/0305-750X(91)90039-
K
Schor, J., Fan, W., Kelly, O., Gu, G., Bezdenezhnykh, T., Bridson-Hubbard, N., 2022.
Assessing global trials of reduced work time with no reduction in pay.
SECO, 2020. Procedures for leave and public holidays [WWW Document]. URL
https://www.kmu.admin.ch/kmu/en/home/concrete-know-how/personnel/employment-
law/working-hours/leave-public-holidays.html (accessed 9.4.23).
Sen, A., 1977. Rational Fools: A Critique of the Behavioral Foundations of Economic
Theory. Philos Public Aff 6, 317–344.
Sensat, J., Constantine, G., 1975. A Critique of the Foundations of Utility Theory. Sci Soc
39, 157–179.
Shao, Q., Rodríguez-Labajos, B., 2016. Does decreasing working time reduce environmental
pressures? New evidence based on dynamic panel approach. J Clean Prod 125, 227–235.
https://doi.org/10.1016/J.JCLEPRO.2016.03.037
Shao, Q., Shen, S., 2017. When reduced working time harms the environment: A panel
threshold analysis for EU-15, 1970–2010. J Clean Prod 147, 319–329.
https://doi.org/10.1016/J.JCLEPRO.2017.01.115
Solomons, A., Allen, F., Smitherman, E., 2023. Full list of UK companies working four day
weeks – with staff on FULL pay. The Sun.
The Conference Board, 2023. Total Economy Database [WWW Document]. URL
https://data-central.conference-board.org/ (accessed 8.11.23).
The Labour Party, 2019. It’s Time for Real Change: The Labour Party Manifesto 2019.
Victor, P.A., 2008. Managing without Growth: Slower by Design, not Disaster. Edward Elgar
Publishing.
Ward, A., Zinni, M.B., Marianna, P., 2018. International productivity gaps: Are labour input
measures comparable? (No. 12), OECD Statistics Working Papers.
Wilting, H.C., Vringer, K., 2010. Carbon and land use accounting from a producer’s and
consumer’s perspective - An empirical examination covering the world. Economic
Systems Research 21, 291–310. https://doi.org/10.1080/09535310903541736
Wong, K., Chan, A.H.S., Ngan, S.C., 2019. The Effect of Long Working Hours and Overtime
on Occupational Health: A Meta-Analysis of Evidence from 1998 to 2018. Int J Environ
Res Public Health 16. https://doi.org/10.3390/IJERPH16122102
72
Appendix A: Expressions for labour demand 𝐿! and capital 𝐾
Differentiating Eq. 2 with respect to labour gives:
Appendix A, Eq. 1
G.
G& (6-2 *8)'!&#! 2/
Setting the first order condition '(
') ( + then gives:
Appendix A, Eq. 2
+ ( 6-2 *8)'!&#! 2/
Solving for & then gives an expression for the quantity of labour demanded by firms under
profit-maximising conditions &+ (Eq. 9):
Appendix A, Eq. 3
&+( B)6-2 *8C"
!'/#"
!
Differentiating Eq. 2 with respect to capital gives:
Appendix A, Eq. 4
G.
G' (*)'!#"&"#! 2 0
Setting the first order condition '(
') ( + then gives:
Appendix A, Eq. 5
+ ( *)'!#"&"#! 2 0
73
Solving for ' then gives an expression for capital under profit-maximising conditions (Eq.
10):
Appendix A, Eq. 6
' ( D "
"#,&E0
*F"
!#"
74
Appendix B: Labour supply function
To maximise utility under the given constraints 4 ( /& and 5 ( 9 2&, a Lagrangian
function is defined in which S% and S& are the Lagrange multipliers corresponding to the
consumption and leisure constraints, respectively.
Appendix B, Eq. 1
T ( 3 @ S%64 2 /&8@ S&65 2 9 @&8
Utility is maximised when the partial derivatives of T with respect to 4 and 5 are equal to
zero. These partial derivatives can be expressed as:
Appendix B, Eq. 2
GT
G4 (G3
G4 @S8
Appendix B, Eq. 3
GT
G5 (G3
G5 @S&
Note that in these expressions, '-
'% and '-
'& are the partial derivatives of the utility function with
respect to consumption and leisure, respectively.
To solve these equations and find the utility-maximising quantities of consumption and
leisure, it is also necessary to formulate an alternative expression for 'ℒ
'& in terms of &, which
appears in both constraints, to remove the Lagrangian multipliers. To this end, the chain rule
can be used as shown:
Appendix B, Eq. 4
GT
G5 (GT
G& AG&
G5
The partial derivative of the Lagrangian with respect to & is given by:
75
Appendix B, Eq. 5
GT
G& ( 2S%/ @ S&
Recalling that ')
'& ( 2-, I use the chain rule and multiply the result in Appendix B, Eq. 5 by
2- to get the alternative expression for 'ℒ
'&:
Appendix B, Eq. 6
GT
G5 ( S%/ 2 S&
Setting these partial derivatives of the Lagrangian with respect to 51and 4 (Appendix B Eqs.
2, 3 and 6) equal to zero and solving the system of equations gives the condition for
maximum utility under the expressed constraints (Eq. 11) :
Appendix B, Eq. 7
G3
G4/ ( G3
G5
Based on the Cobb-Douglas utility function described by Eq. 6, the partial derivative of the
utility function with respect to consumption is given by:
Appendix B, Eq. 8
G3
G4 ( :4$#"5"#$
Similarly, the partial derivative of the utility function with respect to leisure time is given by:
Appendix B, Eq. 9
76
G3
G5 (6-2 :84$5#$
Substituting Appendix B, Eqs. 8 and 9 into Appendix B, Eq. 7 gives the condition for
maximum utility:
Appendix B, Eq. 10
:4$#"5"#$/ ( 6-2 :84$5#$
Cancelling terms and simplifying then gives:
Appendix B, Eq. 11
:5/
4( - 2:
Appendix B, Eq. 11, and its generic form in Appendix B, Eq. 7, represent the conditions under
which households maximise their utility. To find the optimal amount of labour that
households supply to maximise their utility, Appendix B, Eq. 11 can be written in terms of &,
given that 4 ( /& and 5 ( 9 2&.
Appendix B, Eq. 12
:692&8/
/& ( - 2:
Solving for & then gives the quantity of labour households are willing to supply under utility-
maximising conditions &., expressed as a function of the parameter : (Eq. 12) :
Appendix B, Eq. 13
&.(:9
77
Appendix C: Partial derivatives with respect to productivity (𝐴)
Labour (&∗) and leisure time (5∗) at equilibrium depend only on : (see Eqs. 13 and 18), and
as such do not vary in response to changes in ).
The partial derivatives of the equilibrium expressions for consumption (4∗), the ratio of
consumption to leisure time (%∗
&∗), the wage rate (/∗), capital ('∗), and output (%∗) with
respect to productivity ) are given below:
Appendix C, Eq. 1
G4∗
G) (:9)!
"#! E0
*F!
!#"
Appendix C, Eq. 2
GE4∗
5∗F
G) (:E0
*F!
!#" )!
"#!
6-2 :8
Appendix C, Eq. 3
G/∗
G) ( ) !
"#! E0
*F!
!#"
Appendix C, Eq. 4
G'∗
G) (:9E0
*F"
!#" )!
"#!
-2 *
Appendix C, Eq. 5
G%∗
G) (:9E0
*F!
"#! )!
"#!
-2 *
78
Appendix D: Proofs for Eqs. 33 and 34
Proof for Eq. 33
Output at time I is given by Eq. 31:
Appendix D, Eq. 1
%5∗( )5
"
"#!:59E0
*F!
"#!
Given modifications to : and ) by factors J and K, respectively, :56" and )56" can be
expressed as :56" ( J:5 and )56" ( K)5.
Thus, %56"
∗ can be written as:
Appendix D, Eq. 2
%56"
∗(6K)58"
"#!J:59E0
*F!
"#!
Output will fall when %56"
∗, %5∗, which can be expressed using the above equations as:
Appendix D, Eq. 3
6K)58"
"#!J:59E0
*F!
"#! , )5
"
"#!:59E0
*F!
"#!
Cancelling terms then reduces this expression to:
Appendix D, Eq. 4
K"
"#!J , -
Rearranging yields Eq. 33:
Appendix D, Eq. 5
79
J , K "
!#"
Proof for Eq. 34
Emissions at time I is given by Eq. 32:
Appendix D, Eq. 6
>5
∗( ?%:59)5
"
"#!6-2 *8E0
*F!
!#" @ ?&6- 2:589
Given modifications to : and ) by factors J and K, respectively, :56" and )56" can be
expressed as :56" ( J:5 and )56" ( K)5.
Thus, >56"
∗ can be written as:
Appendix D, Eq. 7
>56"
∗( ?%J:596K)58"
"#!6-2 *8E0
*F!
!#" @ ?&6- 2J:589
Emissions will fall when >56"
∗, >5
∗, which can be expressed using the above equations as:
Appendix D, Eq. 8
𝛽!𝜓𝜃"𝑇𝛾 #
#$% 𝐴"
#
#$% (1 − 𝛼),𝑟
𝛼.
%
%$# + 𝛽&(1 − 𝜓𝜃")𝑇 < 𝛽!𝜃"𝑇𝐴"
#
#$% (1 − 𝛼),𝑟
𝛼.
%
%$# + 𝛽&(1 − 𝜃")𝑇
Grouping the consumption and leisure emissions terms and factorising then gives:
Appendix D, Eq. 9
?%:59)5
"
"#!6-2 *8E0
*F!
!#" 6JK "
"#! 2 -8 , ?&:596J2 -8
Dividing both sides by 𝜃"𝑇 reduces the expression to:
80
Appendix D, Eq. 10
?%)5
"
"#!6-2 *8E0
*F!
!#" 6JK "
"#! 2 -8 , ?&6J2-8
Multiplying both sides by −1 then expresses both the LHS and RHS as the differences between
consumption and leisure emissions, respectively, at times 𝑡 and 𝑡 + 1:
Appendix D, Eq. 11
?%)5
"
"#!6-2 *8E0
*F!
!#" Q- 2JK "
"#!R ; ?&6- 2J8
