The Impact of Automation in the Workforce

Page 1

The following paper explores the potential impacts that Connected Autonomous Vehicles (CAVs) could have in the workforce of the UK by 2030. The analysis will present a general introduction to the challenges that automation represents within the Fourth Industrial Revolution. It will provide an overview of current state of the UK regarding CAVs, present a model of the uncertainty factors and its potential impact on the workforce. It will then classify the risk level and conclude with key findings and clear recommendations on how the UK can better prepare to mitigate its potential negative impacts while exploiting the opportunities ahead.

The impact of automation in the workforce An analysis of United Kingdom’s Connected Autonomous Vehicles (CAVs) Industry by 2030 Author: Ana María Saldarriaga Gómez

University College London Deptartment of Science, Engineering and Public Policy January 2018

0


Executive Summary Despite the public narrative where 65% of the U.S. labour market fears robots will replace human jobs, there is not substantial evidence to prove this will be the case. While it is a fact that many jobs will be replaced by automation, (with reliable quantitative forecasts on which ones and by when), all the scenarios indicate that new jobs will be introduced and there is not an imminent threat of social collapse due to this disruption, or at least not until 2030. Taking the industry of CAVs as an example of how automation could impact the labour market of the UK, the analysis shows there are some preventive measurements to be taken to ensure this disruption happens smoothly. The evidence found reveals a deep problem of uncertainty with a multi-actor governance issue. After analysing the current state of play of CAVs within the UK market, and performing an issue map exercise, the uncertainty problem was structured in three core topics: Jobs Disruption, Implementation Timelines and Coordination Capacity. The overall conclusion is that the UK has a strong delivery, analytical and regulatory capacity. Nonetheless, it has a coordination capacity between the Dpt. of Work and Pensions with other departments that need to be addressed urgently. Key findings on how CAVs could potentially impact the UK labour market. Uncertainty Issue Jobs Disruption

Level of Risk Moderate

Implementation Timelines

Low

Key findings • • • •

30% of the UK labour market is at risk of automation (15 M jobs approximately) 950,000+ jobs will be within in the Transport and Logistic Sector CAVs are projected to create 951,000+ new jobs The population at risk is not trained to take over the jobs that will be introduced

In the five stages of automation, human-based decision would still be legally required, implying there is not an imminent risk of sudden job loss Full automation –where no human is needed under no circumstance- is expected to be achieved only after 2030 The UK consumer market showed a decrease in interest in CAVs due to financial privacy and safety concerns. This behaviour can alter the current projections

• •

Coordination Capacity

High

• •

The actors involved in the creation, regulation and commercialization of CAVs technologies do not see themselves as responsible for its potential impacts in the labour market. (Dpt. of Transport, Education, Strategy, Private Sector, etc) The Dpt. of Work and Pensions, the stakeholder responsible to ensure decent jobs, earning power and a stable welfare in the UK, does not seem to have under its radar the potential disruption that automation can have in the workforce, both in employment and in different earning power, automatically affecting the entire pensions system

Due to the level of risk the coordination capacity issue possesses, the conclusions of this research recommend to the Department of Work and Pensions to implement a Performance-Stat Model that allows them to coordinate efforts and maximise resources with the Departments of Transport, Education and Business Strategy. The limitations of the analysis and this model will be identified through the paper. The expected outcomes that the model should cover are: • •

Collect and monitor information regarding how automation is affecting the UK labour market that the Dpt. of Work and Pensions can use to make data-based preventive and reactive decisions to guarantee a smooth implementation of CAVs and other automation technologies. Create an accountability system that synergies the collective effort of the four UK Government Departments to identify and manage potential over and underlaps.

1


Context Every industrial revolution has always raised the same concern: How will the newly introduced technology impact the workforce? This fear is due to the human tendency to resist change. However, as history shows, change is inevitable, and society not only keeps on moving forward but instead raises the living standards for the entire humankind. The last three industrial revolutions show how the question in place is natural, but also unnecessary. In the past, many jobs were indeed replaced, but simultaneously, many more were created (in which most of them could not have been predicted at that time) (Furman 2016). As we are now experiencing the fourth industrial revolution, it is easy to identify how the collective mindset of society is showing the same predicted behaviour. As the think-tank, Pew Research, concludes, 65% of Americans unsurprisingly fear that artificial intelligence will replace human labour, although 80% of them believe that their own job will be safe (Smith 2016). But if the lesson from history teaches us that there should be no reason to concern and that we should, in fact, embrace change, why should it be different this time? At first glance, nothing. But there is one new aspect to the changing environment, and that is the speed of change, and hence, the capacity of society to adapt effectively (Furman 2016). In the past, the impacts of each revolution were seen in a span of centuries, but with this one, the changes are bigger, and the timespan is less than decades (see Appendix A). The world has changed more in the last 15 years than it has ever done in the previous 200, yet today is the slowest pace we will ever experience, as new technologies will keep emerging and the interconnectedness of systems will keep getting stronger (Poynter 2015). This speed of change has brought many positive outcomes such as the increase of productivity - particularly significant in countries with an aging population - but it also entails major challenges due to its effects on the labour market. The world has already experienced the introduction of many automation technologies without a traumatic disruption in the labour market, although the education systems are just starting to catch up to train the upcoming generations with the 21st century demanded skills (Nesta 2017). This low impact is because the world has only seen the basic implementation of Narrowed Artificial Intelligence (AI), where only single and repetitive tasks are prompt to automatization. However, more complex stages of Narrowed AI are being implemented, and the development of General AI (technology that can perform multiple tasks, prioritise, adapt to changing environments, self-learn and make its own decisions) is also coming in the next 10-30 years (Ban 2017). Nevertheless, it is important to highlight it is virtually impossible to predict technology development, and this does not mean there are no reasons to concern regarding Narrowed AI. As the ONS workforce jobs survey for employment reveals, it is expected that automation will replace 30% of current jobs in the UK by 2030 (PwC Blogs 2017). An Economic Outlook report shows how in the United Kingdom, the transportation and storage industry is projected to be the second industry to be automatized after water and waste management (see Appendix B for infographic) However, due to the rapid introduction of CAVs in the market, it is projected to be this is the industry that will experience the highest and most imminent risk when it comes to jobs disruption. (PwC, 2017). This paper aims to study the uncertainty and coordination problems CAVs could potentially represent to the United Kingdom until 2030. Its intended audience are high-level leaders inside the Department of Transport, bureaucrats and policy makers of other departments (i.e. Transport, Education, Business and Strategy, etc.) and the general public with an interest in this topic. The analysis will be done firstly, by understanding the current state of play in the UK regarding CAVs; secondly, by identifying and classifying the type of challenges of how CAVs could potentially affect the workforce; finally, by concluding with key findings that will lead to a set of policy recommendations on how the UK can better prepare for the potential impacts, if any, CAVs will bring into the workforce.

2


State of Play and Policy Problem While the early stages of CAVs are slowly being introduced in numerous countries, there are still significant concerns that could delay the implementation of the more advanced stages. Questions such as “How safe is safe enough to fully implement CAVs in the market?" are in the policy arena and very few people and organizations want to take responsibility (Hook 2017). However, as research shows, the faster CAVs are in the streets, the faster it will be the improvement of its technology, and therefore, increased its benefits, in safety, environmental and productivity aspects (RAND 2017). It is therefore surprising to find that the United Kingdom is one of the few countries that has developed a holistic strategy not only to embrace this technology as a critical pillar of its economy but to become a leading country of technology development and regulation of this field (CCVA-1 2017). Some of the key strategies implemented that have made the UK a leader in the CAV industry are described in Table 1 below: Table 1: UK Government activity timeline towards CAV (Connected Autonomous Vehicles) Type of strategy Regulatory Sandbox and Test-Beds Regulatory strategy of Command and control Regulatory strategy of Regulated self-regulation Public-Private Partnerships

R&D Investment to drive growth and productivity

Outcome Launch of MERIDIAN (CCAV-2 2017): £100m CAV program to create a cluster of excellence in driverless car testing, along the M40 corridor between Coventry and London 4 cities testing driverless car projects: Bristol, Coventry, Greenwich and Milton Keynes Introduction of the Automated and Electric Vehicles Bill 2017-19 Launch the code of practice for testing driverless cars by Department for Transport’s (DfT-1 2015) Establishing the Centre for Connected and Autonomous Vehicles (CCAV) Publishing a call for evidence on the UK’s testing ecosystem for connected and autonomous vehicles Launching a consultation on advanced driver assistance systems and automated vehicle technologies (DfT-2 2016) -Launching the Department for BEIS £20 million feasibility studies and collaborative research and development competition (Innovate UK 2015) -£30 million collaborative research and development and feasibility study competition (Innovate UK 2016) Source: Saldarriaga 2017, adapted from UK Gov CCAV, 2017

Furthermore, the UK government has publicly launched two reports regarding the future of the country regarding CAVs and industrial strategy. The first one is the Market Forecast for Connected Autonomous Vehicles released in July 2017 (CCVA-3 2017), and the second one is the UK Industrial Strategy published in Nov 2017 after 16 months of work (HM Gov 2017). Both reports support much of the evidence presented in this paper, and it provides two key insights. Firstly, the UK Industrial Strategy shows an incredibly strong focus on how CAVs will be a significant driver for national productivity and, it separately addresses other strategies that could mitigate job-loss due to automation, such as education training for 21st challenges. Secondly, the Market Forecast effectively includes a small section on the impact that CAVs may have in the workforce by 2035. However, it analyses its effects from the perspective of the jobs that will be potentially created, but it fails to analyse the negative disruption it may have in the current job market. These two reports combined with the activity timeline showed in Table 1, are the backbone for the policy problem identified in this paper. Together, they conclude that the United Kingdom has taken significant steps to adapt CAVs into the market – and it

3


is a pioneer country in the regulation of this industry – but it is still missing to deeply analyse the uncertainties around the implementation of CAVs within the current job force from the jobs elimination perspective. Moreover, if after studying this uncertainty any adverse impacts is revealed, the UK is missing to develop concrete strategies to mitigate them. This uncertainty problem can be classified as a wicked policy issue. It is somewhat difficult to predict, it evolves under an environment that is continually changing, and ultimately, there is not a perfect solution but rather decisions that can foster “better or worse” outcomes as it will be seen in the recommendations section. An issue map was the framework chosen to structure this policy problem where three main categories of uncertainty were identified (see Figure 1). Each of them shows the main questions that will be addressed in the analysis section of this paper. Figure 1: Policy Issue, three main uncertainties CAVs could have in UK’s workforce

Source: Saldarriaga, 2017

Case Analysis One important consideration to acknowledge is that CAVs are examples of socio-technical environmental systems that present a multi-scale and multi-actor governance issue. Not only CAVs are one of the main pillars of the UK industrial strategy to achieving its transition to ultra-low and zero emission vehicles, but also, its technological development covers all five facets of governance proposed by Gerry Stoker (Stoker 1998). The governance around the impact CAVs may have on jobs has (i) a multi set of institutions and actors that expand beyond government, (ii) there are blurry boundaries and responsibilities within its social and economic issues, (iii) it identifies the power dependency involved in the relationships between institutions engaged in collective action, (iv) it has an autonomous and self-governing network of actors and finally (v) it is proven that the CAV industry can bring results without the need for government to use its authority, but instead steer and guide regulations and strategies around it. The criteria selected to analyse the policy problem presented is a qualitative model (Table 2) derived from the same issue map that revealed the three types of uncertainty. The model was constructed based on the Uncertainty Matrix (Cleden 2009) presented in detailed in Appendix C (Kim 2012) with the aim to make the analysis homogeneous and therefore, make an accurate assessment on the prioritisation of the policy recommendations. Table 2: Qualitative classification of risk, based on selected criteria to measure uncertainty. Risk Classification High Medium Low

Uncertainty Criteria Studies performed

Actions taken/planned

Adaptation feasibility

-Are there any formal and reliable studies that show the current state of the issue? -Are there any reliable call to actions?

-Are there clear policies, roadmaps, strategies or partnerships that show how the intended action steps in the short and in the long term?

-Is there enough time to implement preventing and reactive measurements? -Are there enough resources (money, time, H.R. expertise, etc) to foster adaption? Source: Saldarriaga 2017

4


1. Jobs Disruption Analysis The two main reasons the UK is highly committed to implementing CAVs in the streets are citizen safety and market opportunity. Considering that 96.2% of car accidents are caused by human errors that could easily be prevented with automation (i.e. drunk driving, texting, tiredness, etc.), more than 26,000 accidents could be avoided saving more than 2,500 lives per year (Smith 2013). And as if this was not a strong enough reason for any government to start investing in CAVs, the UK is already expecting that in 2035, CAVs will add £28-52bn billion per year to the economy (CCVA-3 2017). Because of these two reasons, wondering if the CAV industry will disrupt the workforce of the UK is the wrong question. The right questions should be in the lines of which jobs and to which degree. The 2017 UK Market Forecast Report estimates that the labour intensity of producing CAV hardware is 5-10 jobs per £1 million output and the labour intensity of producing CAVrelated software is 5-8 jobs per £1 million output. With this logic, the report makes a throughout analysis of the type of jobs that will be created (listed in column A of Table 3) and the expected number. The conclusion is that 951,000+ new jobs (151,000 direct and 800,000 indirect) will be introduced to the UK automotive industry by 2035. At the same time, the UK Economic Outlook performed by PwC (PwC 2017) concluded that 30% of the UK labour market is at risk of automation (15 M jobs approximately), and that there are an estimated of 950,000+ jobs that will be automated in the Transport and Logistic Sector (listed in column B of Table 3). Table 3: Types of jobs with tendency of automation and creation by 2030 A) Types of jobs with potential of creation (951.000+ approx.) HMI software Data security software Connectivity / V2X software Mapping & path planning software Control systems software Safety-related HMI hardware Control systems and computing hardware Connectivity hardware Sensor-supporting hardware Sensing & local mapping hardware

B) Type of jobs with potential of automatization (950.000+ approx.) Truck/Logistics Drivers Bus/Train/Tram Drivers Taxi/Uber/Personal drivers School drivers Delivery drivers Ambulance drivers Valet and parking service jobs Extended Scope: -Health industry due to fewer accidents -Business that cater to drivers (i.e. restaurants, motels, etc.)

Source: Saldarriaga 2017, extracted from PwC UK Economic Outlook Report and 2017 UK CAV Market Forecast Report

It is a striking coincidence that the ratio between both, potentially automated and created jobs, is virtually 1:1. However, this does not mean there are no reasons for concern. Although the PwC report reaffirms that that automation will boost productivity and create new employment opportunities, it is also clear at making an urgent call to action to prevent the increasing gap in inequality that artificial intelligence could have in the low-skilled employees (PwC 2017). This call to action becomes even more relevant, especially considering that the CAVs Market Report fails to analyse the impact of jobs from the perspective of job elimination due to automation. Hence, it also neglects to admit that the profiles that are at risk, do not have the technical training (nor any other transferable skill) to perform the jobs that are prompt to be introduced with automation. Nevertheless, as planned in its Industrial Strategy, three critical education policies could potentially lessen this risk. The first one is to transform the technical and higher education system with a strong focus on the younger generations. The second one is to reduce the existing talent shortage in STEM (Science Engineering and Mathematics) profiles by investing £406M in training programs. Finally, creating a national retraining scheme to help mid-career employees to re-skill in the digital and construction areas starting with a £64M investment (HM Gov 2017). One limitation of this analysis is that it does not take into consideration the threats or opportunities that Brexit could represent in both, potentially automated and created jobs However, it is possible to affirm that it is highly likely that there will be an impact at least in the reallocation of talents (WEF 2016). There are at least 60,000 Eastern European truck drivers currently

5


employed in the UK, where the logistic market is at current shortage of 35,000 drivers, and it is expected to keep growing (AFT 2017). Furthermore, UK is expected to need 1.8 million STEM profiles by 2025, and it currently relies on foreign talents to barely keep up with the market demands (Engineering UK 2017). Because of the aforementioned analysis, it is possible to conclude a) the official studies of CAVs impact on jobs do exist, but are incomplete and b) there are no formal action steps taken nor strategies in the future to mitigate its potential negative consequences. These two conclusions represent a threat mainly because of its lack of monitoring, implying an awareness gap from the Government’s side. However, considering that the most feasible solution to tackle this issue would be a comprehensive re-skill strategy for the population with jobs at risk, it is possible to see how the education reform and training programs proposed in Industrial Strategy plan, could be easily aligned to mitigate this issue. Therefore, considering all these insights, it possible to classify the risk uncertainty of jobs disruption as medium.

2. Implementation Timelines Analysis In all likelihood, while automation will not replace all jobs in the automotive industry, but in fact, will generate new ones, there will be a period of transition that may disproportionately affect low-skilled employees such as drivers (RSA 2017). This period of turnover will not happen over time and it is currently divided in four different stages explain in Table 4. Table 4: Levels of driving automation as defined by SAE International J3016 Stage 0: No Automation

Stage 1: Driver Assistance

Stage 2: Partial Automation

Human driver performs part or all the dynamic driving task; in particular, the driver is responsible for monitoring the environment and any action taken by the automation system Human driver performs all aspects of dynamic driving tasks

System can perform either steering or acceleration (e.g. Park Assist, Adaptive Cruise Control)

System can perform both steering and acceleration (e.g. Traffic Jam Assist)

Stage 3: Conditional Automation

Stage 4: High Automation

Stage 5: Full automation

System performs entire dynamic driving task while engaged, including monitoring and response as well as steering and acceleration

Human driver may be requested to intervene (fall-back) (e.g. Intersection Pilot, Platooning)

Full automation in some driving modes (e.g. Urban automated driving)

Full automation in all driving modes

Source: Adapted from SAE International J3016 taxonomy and definitions (full diagram shown in Appendix D).

Appendix E shows a 2017 Deloitte Insights publication that shows the timeline projection of the implementation of these five stages into the global market. Figure 3 shows a visual summary of the timeline within the phases. The current jobs at risk are not going to present any disruptions until phase 4, because a human back up would still be legally required in case of emergencies or unusual situations that are yet to be discovered the challenge would begin after a considerable stabilization of the phase 4 (Deloitte Insights 2017). It is therefore expected, once the market has had first-hands experience with the technology and has gain an adequate degree of trust in it, institutions such as law makers and insurance companies would be willing to consider to fully roll out driverless cars without a human back up. Figure 3: Simplified timeline of CAVs implementation in the UK market (Appendix E)

There are, nevertheless, some limitations to this analysis that challenge the reliability of this timeline. Firstly, technological development is unpredictable and when it comes to artificial intelligence, there have been significant margin errors Âą20-30%

6


both in accelerating and delaying predicted timelines impact through history (Deloitte Insights 2017). Secondly, regulations and insurance policies could hinder the speed of implementation as most institutions do not want to take full responsibilities of the negative consequences that may come alone with the technology (Baumann 2017). Finally, there is contradictory data regarding the potential acceptance to this technology within the general public. While there is an exponential growth in the number of searches requests that google has had since 2004 to 2005 a global Global Automotive Consumer Survey shows that in the UK, there has been a massive drop in interest of driverless car (see Appendix F) (Deloitte UK, 2017). The main reason is financial unaffordability based on a consumer perception survey Nevertheless, removing the fun of driving, along with privacy and safety concerns were also at the top of the list. (See infographic Appendix H) (Danise, 2016). Considering there are recent and reliable studies projecting the chronological implementation of CAVs aligned with the stages of automation and clearly stating the need for human job, it is safe to say that –considering the limitations- there is little uncertainty in this issue. In the event the timelines accelerate, human labor it is still likely to be legally required until 2030, and, in if it delays, it would just give more time to the workers to adapt. Furthermore, the UK has already taken significant steps to invest in the technology development (R&D) and enable the regulation (bills and test beds) fostering a smooth execution of the timelines proposed. Lastly, despite the decrease in interest in the UK consumer market, the feasibility of its adaptation is high considering that infrastructure investments that are coming, are expected to remove the concerns of the market. Hence, the risk uncertainty in this is issue is classified as low.

3. Coordination Capacity Analysis The research of this policy issue has revealed the UK has a strong administrative capacity to implement CAVs into the market. This paper has compiled some examples (refer to Table 1) that proves the government's readiness for delivery, analytical and regulatory capacity. However, as stated in the Job Disruption Analysis Section, it seems that there could be a coordination gap between different governmental departments, that implies lack of awareness on how CAVs could impact the workforce. As a first step to test this assumption, a stakeholder mapping analysis was performed and where 13 actor groups have been identified. Figure 2 shows the stakeholder matrix outcome and the classification of each depending on their level of influence vs degree of impact. Figure 2: Stakeholder Matrix

The impact that CAVs may have in the workforce would only affect the stakeholder of the fourth quadrant (high impact, low influence), as for the rest, they are expected to steer, adapt or both. Despite not being the intended audience for this analysis, some coordination issues were spotted among the private sector (automobile, AI technology and Insurance companies). Problems of patents and liabilities are currently slowing down the timelines projected in Figure 3; However, its direct relation on how this issue could have a level of influence on the impact of jobs was not found to be significant, or at least not when compared with the coordination issues found in the public sector.

7


There are four government departments identified as key actors for the impact of automation of CAVs in the workforce. The Department of Transport, of Business, Energy and Industrial Strategy, of Education and finally, of Work and Pensions. The first two seem to take the lead on this policy issue, considering they are both founders of the Centre for Connected Autonomous Vehicles (CCAV) –the centre that conducted the market forecast responsible for studying the impact of CAVs in the workforce. However, after interviewing Greg Clark, Board Member of the Transport of London, regarding his thoughts on why the CCAV had left out the negative impacts of CAVs in the work labour, his answer provided an honest insight. He affirmed that, while the Dpt. of Transport is proud to employ approximately 5,000 drivers on its different systems, being an employer is not their ultimate objective. Their job is to transport citizens from A to B, in the most efficient, affordable and sustainable way possible. (Clark 2017) This insight opened the door to identify the stakeholder(s) whose objective was to protect the interest of these workers, and that is how the Departament of Work and Pensions came as the primary responsible (UK Gov 2017). The uncertainties studies applied to this actor followed the model proposed in Table 2. It aimed to identify A) if the Dpt. of Work had had any involvement in the developing the study of the CAV Market Forecast and its impacts on jobs, B) if the Dpt. had raised any concerns regarding artificial intelligence and C) if the Dpt. had taken any action plan or preventive measures to mitigate its potential negative impacts on the workforce. The research did not show their involvement in the creation of the CAV Market Forecasts nor any official strategy to mitigate the risk. Nevertheless, the fourth actor in this analysis –The Department of Education- has already taken the lead on this issue as found in the Industrial Strategy Report previously explained. Hence, a basic alignment of common objectives could help to create a coordination strategy to tackle this issue together. Regarding questions B and C, the research showed that The Department of Work has indeed taken preventive measurements regarding the impact of AI in the workforce, but only from the pension and wages system perspective (Stweart and Mason 2017). This is particularly important because when analysing from a holistic perspective the impact that automation could have in the workforce the growing population trend combined with a prolonged life expectancy are two additional threats to the policy issue of this paper (see causal loop diagram in Appendix G). In July 2017, the Dpt. Of Work announced the highly criticised decision to increase the pension age from 67 to 68 starting in 2037 due to the impact that technology was having in the life expectancy. However, it does not acknowledge the threat if technology dramatically disrupts the labour market, this will automatically disrupt the pension and welfare system, making it a pressing priority for them. This disruption can be both an opportunity and a threat to the job market. In its simplest form, Dhaval Joshi, an economist from BCA research explains both scenarios showing how the Moravec’s paradox will have a significant impact on the workforce. The paradox is, while robots find simple the activities that human consider difficult, they find difficult the activities humans consider simple. This correlation is important because it highlights the opportunity in which automation will create jobs and that humans will not be replaced, but complemented. However, it also implies there will be a shift in the type of jobs humans would perform, and this may decrease the quality of those jobs, negatively impacting the earning power and pension systems (Elliot 2017). This theory is also reaffirmed by a report of UK Trade Union Congress, which states that the productivity that automation entails could help in reducing the pension age. However, for this to happen, he makes an open call to debate on who benefits from this rising prosperity and how workers can get a fair share, debates that until now have not taken place (TUC 2017). There were two limitations in this process. Firstly, there was an attempt to perform an interview with different representatives at different levels inside the Department of Work and Pensions with the intention to validate the accuracy of these findings. Though the meetings could not take place, they are highly recommended for further analysis to improve the rigour of the investigation and to compensate the lack of quantitative evidence. Secondly, most of the information found regarding the impact of automation in the UK workforce and its relation with the Department of Work and Pensions was from a general automation perspective. Finding specific information about the CAVs industry was not possible.

8


The research on this issue concludes, although there are several theories and expert’s opinions –mainly gathered by the mainstream media- there are little to none formal studies on how automation will impact the UK welfare and pension systems. It also concludes that the Department of Work and Pensions, not only does not have a clear stand, but it seems to be unaware of this issue. This conclusion is validated when analysing the report that the National Audit Office performed on the activities of the department, no concrete actions nor developing of future strategies were found (Redaway et al. 2017). Considering all the above findings, the risk uncertainty for this issue is classified as high.

Case Findings and Conclusions Despite the public narrative where 65% of the U.S. labour market fears robots will replace human jobs, there is not substantial evidence to prove this will be the case (RAND-2 2016). While it is a fact that many jobs will be replaced by automation, (with reliable quantitative forecasts on which ones and by when –PwC, 2016), all the scenarios indicate that new jobs will be introduced and there is not an imminent threat of social collapse due to this disruption, or at least not until 2030. Taking the industry of CAVs as an example of how automation could impact the labour market of the UK, the analysis shows there are some preventive measurements to be taken to ensure this disruption happens smoothly. The evidence gathered reveals a deep problem of uncertainty with a multi-actor governance issue. After analysing the current state of play of CAVs within the UK market, and performing an issue map exercise, the uncertainty problem was categorised in three core topics: Jobs Disruption, Implementation Timelines and Coordination Capacity. The overall conclusion is that the UK has a strong delivery, analytical and regulatory capacity. Nonetheless, it has a coordination capacity between the Department of Work and Pensions with other Departments (Transport, Education and Business and Industrial Strategy) that need to be addressed urgently. The key findings on how CAVs could impact the UK labour market are summarised in Table 5. Table 5: Case findings, Impact of CAVs within the UK labour Market Uncertainty Issue Jobs Disruption

Level of Risk Moderate

Key findings • • • • •

Implementation Timelines

Low

• • • • •

30% of the UK labour market is at risk of automation (15 M jobs approximately) 950,000+ jobs will be within in the Transport and Logistic Sector CAVs are projected to create 951,000+ new jobs (151,000 direct and 800,000 indirect) The profiles of the jobs that will be created does not match with the profiles of the population at risk The Dpt. of Education has already a plan to re-skill the population with technical skills, however, it is not directly connected with the CAVs industry nor it was cocreated with the Dpt. of Work and Pensions. There are five stages of automations, that help classifying the degree of automation and therefore, the risk. In all automation stages, human-based decision would still be legally required, implying there is not an imminent risk of sudden job loss. Full automation –where no human is needed under no circumstance- is expected to be achieved only after 2030. This gives a timeframe for the population at risk to reskill and reallocate in other available jobs. It is more likely that the timeline would be delayed than accelerated, but in neither scenario, the level of risk changes. Unexpected market reactions, such as the decreased of interest in CAVs in the UK population due to financial privacy and safety concerns, could generate unexpected delays and changes in the projections.

9


Coordination Capacity

High

• •

•

The actors involved in the creation, regulation and commercialization of CAVs technologies do not see themselves as responsible for its potential impacts in the labour market. The Dpt. of Work and Pensions, the stakeholder responsible to ensure decent jobs, earning power and a stable welfare in the UK, does not seem to have under its radar the potential disruption that automation can have in the workforce, both in employment and in different earning power, automatically affecting the entire pensions system. There are separated efforts to ensure CAVs are successfully implemented in the UK. However, there is a coordination problem between the Dpt. of Work and pensions and the Dpts. of Education, Transport and Business and Ind. Strategy.

Policy Recommendations The conclusions of this research suggest that the coordination capacity should be the issue to be addressed above any other. Not only it possesses the highest risk, but if a proper solution is implemented to solve this problem, there are high chances that the other two issues (job disruption and timeline implementation) would be simultaneously addressed too. Furthermore, it is an issue that only the Department of Work and Labour could take the lead on, not just because is under its level of responsibilities, but also, because it is the largest government department in the UK with a higher potential to access to resources (i.e. HR). Considering the coordination gap is happening between four public agencies with blurry boundaries of responsibility regarding the impact of CAVs (and other automation technologies) in the workforce, we recommend implementing a boundary object analysis such as a Performance-Stat Model (Behn, 2007). The Performance-Stat Model is a leadership strategy to mobilise public agencies to produce specific results. While it is mainly composed of highly operational tasks such as periodic meetings, follow up of targets, and monitoring of data, the statistics it gathers become a strategic resource that enables the leadership body (i.e. the Secretary of State for the Work and Pensions Department) to identify past trends and predict future behaviours that enables to make accurate and timely decisions. There are two primary objectives of how this model could help in managing the impact of CAVs in the UK labour market, and they are based on detection and prediction of this issue. The expected outcomes of these two objectives would be to: A) Collect and monitor information regarding how automation is affecting the UK labour market B) Create an accountability system that encourages the Department of Work and Pensions to take timely preventive and reactive decisions, ensuring a smooth implementation of CAVs and other automation technologies C) Foster a collective environment that enable synergies across all four departments that helps to identify and manage potential over and underlaps. Considering this is a wicked policy issue, in which there are no black and white situations nor quick solutions, there are some limitations on how this model can be applied. For example, the qualitative and public value narratives around our policy problem could be hard to integrate inside a statistical modulation. Furthermore, although this model is reasonably affordable to implement, it has proven to be time-consuming, therefore, aligning an organizational culture around its importance across the four departments, could be a full-time job for whoever would be the responsible to implement it. Nonetheless, the findings of this analysis and its recommendation were shared during an in-person interview with Benjamin Kulka, a senior researcher of the Centre for London expert in this field, and his overall remarks validate the relevance, applicability and sense of urgency of the recommended course of action (Kulka, 2017).

10


Appendices Appendix A: The four industrial revolutions, timeline and characteristics

Copyright Š 1998-2017, Dr. Jean-Paul Rodriguez, Dept. of Global Studies & Geography , Hofstra University, New York, USA

11


Appendix B: Infographic, proportion of jobs facing potential high risks of automation

Guardian graphic | Source: NS workforce jobs survey for employment shares (2016); PwC estimates for last column using PIAAC data from OECD. High risk of automation is defined as 70 or over based on technical feasibility considerations only

12


Appendix C: Uncertainty Matrix, The impact of CAVs in the UK labour Market

Appendix H: Uncertainty Matrix, The impact of CAVs in the UK labour Market

Source: 2015 Survey, NerdWallet

13


Appendix D: SAE Summary of levels of driving automation. DDT = dynamic driving task; OEDR = object and event detection and response; ODD = operational domain design; ADS = automated driving system.1

1

SAE International, 2016. J3016TM. Surface vehicle recommended practice – Taxonomy and definitions for terms relating to driving automation systems for on-road motor vehicles

14


Appendix E: Driverless Technologies Predicted Timeline

15


Appendix F: Worldwide interest vs UK consumer interest in driverless cars

Source: GoogleTrends, July 2017

Source: 2016 and 2014 Global Automotive Consumer Survey, Deloitte

16


Appendix G: Causal Loop Diagram: The impact of automation in the workforce.

Source: Saldarriaga, 2017

17


Bibliography 1.

Furman J. (2016) Is this time different? The opportunities and challenges of AI [online] Available at: https://obamawhitehouse.archives.gov/sites/default/files/page/files/20160707_cea_ai_furman.pdf [Accessed 2 Jan. 2018].

2.

Smith, A. (2016). Public Predictions for the Future of Workforce Automation. [online] Pew Research Center: Internet, Science & Tech. Available at: https://www.linkedin.com/pulse/rate-change-today-slowest-you-ever-experience-ray-poynter// [Accessed 2 Jan. 2018]

3.

Poynter, R. (2015) – The rate of change today is the slowest you will ever experience. [Online] Available at: https://ourworldindata.org/technological-progress/ [Accessed 2 Jan. 2018].

4.

NESTA, (2017). - Creativity vs. Robots [online] Available at: https://www.nesta.org.uk/sites/default/files/creativity_vs._robots_wv.pdf [Accessed 7 Dec. 2017].

5.

Ban, Y. (2017). Types of Artificial Intelligence. [online] Futuretimeline.net. Available at: http://www.futuretimeline.net/blog/2017/02/13-2.htm [Accessed 20 Dec. 2018].

6.

Pwc Blogs (2017). Up to 30% of existing UK jobs could be impacted by automation by early 2030s, but this should be offset by job gains elsewhere in economy - Press room. [online] Available at: http://pwc.blogs.com/press_room/2017/03/up-to-30of-existing-uk-jobs-could-be-impacted-by-automation-by-early-2030s-but-this-should-be-offse.html [Accessed 9 Dec. 2018].

7.

PwC. (2017). UK Economic Outlook. [online] Available at: https://www.pwc.co.uk/economic-services/ukeo/pwc-uk-economicoutlook-full-report-july-2017.pdf [Accessed 6 Jan. 2018].

8.

Hook, L. (2018). For driverless cars, how safe is safe enough? [online] Financial Times. Available at: https://www.ft.com/content/70924ace-cf0d-11e7-b781-794ce08b24dc [Accessed 5 Jan. 2018].

9.

RAND Corp (2017). Why Waiting for Perfect Autonomous Vehicles May Cost Lives. [online] Available at: https://www.rand.org/blog/articles/2017/11/why-waiting-for-perfect-autonomous-vehicles-may-cost-lives.html [Accessed 7 Jan. 2018].

10. CCVA-1 Gov.uk. (2017). Centre for Connected Automated Vehicles research and development projects 2017 [online] Available at: https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/650444/ccav-research-and-developmentprojects-2017.pdf [Accessed Nov 30. 2017]. 11. CCVA-2 Gov.uk. (2017). Government launches MERIDIAN to accelerate connected autonomous vehicle technology development in the UK - GOV.UK . [online] Available at: https://www.gov.uk/government/news/government-launchesmeridian-to-accelerate-connected-autonomous-vehicle-technology-development-in-the-uk [Accessed Nov 30. 2017]. 12. DfT-1 Gov.uk. (2018). Testing automated vehicle technologies in public - GOV.UK . [online] Available at: https://www.gov.uk/government/publications/automated-vehicle-technologies-testing-code-of-practice [Accessed Nov 30. 2017]. 13. DfT-2 Gov.uk. (2016). Using advanced driver assistance systems and automated vehicle technologies - GOV.UK . [online] Available at: https://www.gov.uk/government/consultations/advanced-driver-assistance-systems-and-automated-vehicletechnologies-supporting-their-use-in-the-uk [Accessed Nov 30. 2017]. 14. Innovate UK (2015). Connected and autonomous vehicles - Funding competition - innovateuk. [online] Available at: https://interact.innovateuk.org/competition-display-page/-/asset_publisher/RqEt2AKmEBhi/content/connected-andautonomous-vehicles [Accessed Nov 30. 2017]. 15. Innovate UK (2016). Connected and autonomous vehicles: apply for business funding - GOV.UK . [online] Available at: https://www.gov.uk/government/news/connected-and-autonomous-vehicles-apply-for-business-funding [Accessed Nov 30. 2017].

18


16. HM Gov. (2017). Industrial Strategy: building a Britain fit for the future [online] Available at: https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/664563/industrial-strategy-white-paper-webready-version.pdf [Accessed 8 Dec. 2017]. 17. CCVA-3 Gov.uk. (2017). Connected and autonomous vehicles: Market forecast - GOV.UK . [online] Available at: https://www.gov.uk/government/publications/connected-and-autonomous-vehicles-market-forcecast [Accessed 7 Dec. 2017]. 18. Stoker, G. (1998). Governance as theory: Five prepositions. [online] Available at: http://catedras.fsoc.uba.ar/rusailh/Unidad%201/Stoker%202002,%20Governance%20as%20theory,%20five%20propositions. pdf [Accessed 7 Jan. 2018]. 19. Cleden, D. (2009). Managing project uncertainty. Farnham, UK: Gower. 20. Kim, S. D. (2012). Characterizing unknown unknowns. Paper presented at PMI® Global Congress 2012—North America, Vancouver, British Columbia, Canada. Newtown Square, PA: Project Management Institute. 21. Smith, B. Cyberlaw.stanford.edu. (2013). Human error as a cause of vehicle crashes. [online] Available at: http://cyberlaw.stanford.edu/blog/2013/12/human-error-cause-vehicle-crashes [Accessed 3 Jan. 2018]. 22. WEF (2016) World Economic Forum, Future of Jobs Report 2016. [online] Available at: http://www3.weforum.org/docs/WEF_Future_of_Jobs.pdf [Accessed 7 Dec. 2017]. 23. AFT (2017). Driver Shortage Report [online] Available at: http://www.repgraph.co.uk/files/reports/The%20driver%20shortage%20report.pdf [Accessed 8 Dec. 2017]. 24. Engineering UK (2017). Synopsys and Recommendations [online] engineeringuk.com Available at: https://www.engineeringuk.com/media/1356/enguk_report_2017_synopsis.pdf [Accessed 3 Jan. 2018]. 25. RSA (2017). The age of automation: Artificial Intelligence, robotics and the future of low-skilled work - RSA. [online] Available at: https://www.thersa.org/discover/publications-and-articles/reports/the-age-ofautomation?utm_medium=referral&utm_source=Guardian&utm_campaign=age-of-automation&utm_content=report# [Accessed 7 Dec. 2017]. 26. Kim, S. D. (2012). Characterizing unknown unknowns. Paper presented at PMI® Global Congress 2012—North America, Vancouver, British Columbia, Canada. Newtown Square, PA: Project Management Institute. 27. Deloitte Insights (2017). Development of self-driving vehicles in the United Kingdom. [online] Available at: https://www2.deloitte.com/insights/us/en/focus/future-of-mobility/development-of-driverless-vehicles-in-the-unitedkingdom.html [Accessed 6 Jan. 2018] 28. Baumann, M. (2017). Why Waiting for Perfect Autonomous Vehicles May Cost Lives. [online] Available at: https://www.rand.org/blog/articles/2017/11/why-waiting-for-perfect-autonomous-vehicles-may-cost-lives.html [Accessed 8 Jan. 2018]. 29. Deloitte UK. (2017). 2017 UK Automotive Consumer Study | Deloitte UK. [online] Available at: https://www2.deloitte.com/uk/en/pages/manufacturing/articles/automotive-consumer-study.html [Accessed 1 Jan. 2018]. 30. Danise, A. (2015). Women Say No Thanks to Driverless Cars, Survey Finds; Men Say Tell Me More - NerdWallet. [online] NerdWallet. Available at: https://www.nerdwallet.com/blog/insurance/survey-consumer-fears-self-driving-cars/ [Accessed 5 Jan. 2018]. 31. Clark, G (2017). Interviewed by: Saldarriaga, Ana during UCL Seminar. [22 Nov. 2017] 32. UK Gov (2017). About us - Department for Work and Pensions - GOV.UK . [online] Available at: https://www.gov.uk/government/organisations/department-for-work-pensions/about [Accessed 3 Jan. 2018]. 33. Stewart, H. and Mason, R. (2017). State pension age to increase seven years earlier than planned. [online] the Guardian. Available at: https://www.theguardian.com/money/2017/jul/19/state-pension-age-to-increase-seven-years-earlier-thanplanned [Accessed 9 Jan. 2018]. 34. Elliott, L. (2017). Robots will not lead to fewer jobs – but the hollowing out of the middle class | Larry Elliott. [online] the Guardian. Available at: https://www.theguardian.com/business/2017/aug/20/robots-are-not-destroying-jobs-but-they-arehollow-out-the-middle-class [Accessed 3 Jan. 2018].

19


35. TUC (2017). Productivity Report –No puzzle about it Tuc.org.uk. (2018). [online] Available at: https://www.tuc.org.uk/sites/default/files/productivitypuzzle.pdf [Accessed 2 Jan. 2018]. 36. Redaway et al. (2017), National Audit Office, A short guide to the Department of Work and Pensions [online] Available at: https://www.nao.org.uk/wp-content/uploads/2017/09/11559-001-DWP-SG_6DP_final.pdf [Accessed 4 Jan. 2018]. 37. RAND-2 (2016). Self-Driving Vehicles Offer Potential Benefits, Policy Challenges for Lawmakers. [online] Available at: https://www.rand.org/pubs/research_reports/RR443-2.html [Accessed 8 Dec. 2017]. 38. Behn, R. (2007). What all majors would like to know about Baltimore’s CitiStats Performance Model. JFK School of Government, Harvard University [online] Available at: http://openeyecommunications.typepad.com/Uploads/CitiStat.pdf [Accessed 8 Dec. 2017]. 39. Kulka, B. (2017). Interviewed by: Saldarriaga, Ana at Centre for London Offices. [10 Oct. 2017]

Further reading: 40. Clegg, R. (2017). UK labour market - Office for National Statistics. [online] Ons.gov.uk. Available at: https://www.ons.gov.uk/employmentandlabourmarket/peopleinwork/employmentandemployeetypes/bulletins/uklabourmarket /november2017#unemployment [Accessed 8 Dec. 2017]. 41. SMMT. (2017). 2017 Automotive Sustainability Report - SMMT. [online] Available at: https://www.smmt.co.uk/industrytopics/sustainability/ [Accessed 8 Dec. 2017]. 42. World Economic Forum. (2017). 4 predictions for the future of work. [online] Available at: https://www.weforum.org/agenda/2017/12/predictions-for-freelance-work-education [Accessed 7 Dec. 2017]. 43. Furman, J. (2016). How to Protect Workers From Job-Stealing Robots. [online] The Atlantic. Available at: https://www.theatlantic.com/business/archive/2016/09/jason-furman-ai/499682/ [Accessed 7 Dec. 2017]. 44. Alles, M. Int J Discl Gov (2009) 6: 85. Available at: https://doi.org/10.1057/jdg.2009.2 45. Weller, C. (2017). Obama just warned Congress about robots taking over jobs that pay less than $20 an hour. [online] Business Insider. Available at: http://uk.businessinsider.com/obama-warns-congress-about-robot-job-takeover-20163?r=US&IR=T [Accessed 7 Dec. 2017]. 46. Business Insider UK 2016, 1. (2017). 10 million self-driving cars will be on the road by 2020. [online] Business Insider. Available at: http://uk.businessinsider.com/report-10-million-self-driving-cars-will-be-on-the-road-by-2020-2015-56?r=US&IR=T [Accessed 8 Dec. 2017]. 47. Obamawhitehouse.archives.gov. (2017). [online] Available at: https://obamawhitehouse.archives.gov/sites/whitehouse.gov/files/documents/Artificial-Intelligence-Automation-Economy.PDF [Accessed 8 Dec. 2017]. 48. Bourgon, J. (2011). A New Synthesis of Public Administration: Serving in the 21st Century. McGill-Queen's University Press. 49. Litman, T. (2018). Autonomous Vehicle Implementation Predictions Implications for Transport Planning - Victoria Transport Policy Institute [online] Available at: http://www.vtpi.org/avip.pdf [Accessed 9 Jan. 2018].

20


Turn static files into dynamic content formats.

Create a flipbook
Issuu converts static files into: digital portfolios, online yearbooks, online catalogs, digital photo albums and more. Sign up and create your flipbook.