WEF FOW Reskilling Revolution

Transcript

1 Insight Report Towa r d s a Reskilling Revolution A Future of Jobs for All In collaboration with The Boston Consulting Group January 2018

2 TERMS OF USE AND DISCLAIMER (herein: Towards a Reskilling Revolution: A Future of Jobs for All “report”) presents information and data that were compiled and/or collected by the World Economic Forum (all information and data referred herein as “data”). Data in this report is subject to change without notice. as used in this report do not in all cases nation and country The terms refer to a territorial entity that is a state as understood by international law and practice. The term covers well-defined, geographically self- contained economic areas that may not be states but for which statistical data are maintained on a separate and independent basis. Although the World Economic Forum takes every reasonable step to ensure that the data thus compiled and/or collected is accurately reflected in this report, the World Economic Forum, its agents, officers, and employees: (i) provide the data “as is, as available” and without warranty of any kind, either express or implied, including, without limitation, warranties of merchantability, fitness for a particular purpose and non-infringement; (ii) make no representations, express or implied, as to the accuracy of the data contained in this report or its suitability for any particular purpose; (iii) accept no liability for any use of the said data or reliance placed on it, in particular, for any interpretation, decisions, or actions based on the data in this report. Other parties may have ownership interests in some of the data contained in this report. The World Economic Forum in no way represents or warrants that it owns or controls all rights in all data, and the World Economic Forum will not be liable to users for any claims brought against users by third parties in connection with their use of any data. The World Economic Forum, its agents, officers, and employees do not endorse or in any respect warrant any third-party products or services by virtue of any data, material, or content referred to or included in this report. Users shall not infringe upon the integrity of the data and in particular shall refrain from any act of alteration of the data that intentionally affects its nature or accuracy. If the data is materially transformed by the user, this must be stated explicitly along with the required source citation. For data compiled by parties other than the World Economic Forum, users must refer to these parties’ terms of use, in particular concerning the attribution, distribution, and reproduction of the data. When data for which the World Economic Forum is the source (herein: “World Economic Forum”) is distributed or reproduced, it must appear accurately and be attributed to the World Economic Forum. This source attribution requirement is attached to any use of data, whether obtained directly from the World Economic Forum or from a user. Users who make World Economic Forum data available to other users through any type of distribution or download environment agree to make reasonable efforts to communicate and promote compliance by their end users with these terms. Users who intend to sell World Economic Forum data as part of a database or as a standalone product must first obtain the permission from the World Economic Forum ([email protected]). World Economic Forum 91-93 route de la Capite CH-1223 Cologny/Geneva Switzerland Tel.: +41 (0)22 869 1212 Fax: +41 (0)22 786 2744 Email: [email protected] www.weforum.org World Economic Forum® © 2018 – All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, including photocopying and recording, or by any information storage and retrieval system. REF 220118

3 Contents 01 Preface Introduction 03 04 Mapping Job Transition Opportunities Is the job transition viable? Is the job transition desirable? 07 Finding Job Transition Pathways for All Leadership lens Individual lens 17 Conclusion 21 Appendix A: Data and Methodology 27 Appendix B: Job Transition Pathways 35 System Initiative Partners 37 Acknowledgements

4

5 Preface KLAUS SCHWAB Founder and Executive Chairman, World Economic Forum As the types of skills needed in the labour market change rapidly, In assessing reskilling pathways and job transition individual workers will have to engage in life-long learning if they opportunities in such detail and at such scale, we aim to move are to remain not just employable but are to achieve fulfilling and the debate on the future of work to new—and practical—territory. rewarding careers that allow them to maximize their employment This report is a beginning. In subsequent publications, the opportunities. For companies, reskilling and upskilling strategies methodology will be extended to include additional perspectives and geographies and applied in collaboration with government will be critical if they are to find the talent they need and to contribute to socially responsible approaches to the future of and business stakeholders to support workers. We also hope work. For policy-makers, reskilling and retraining the existing it inspires similar efforts to think practically yet holistically about managing reskilling, upskilling and job transitions. workforce are essential levers to fuel future economic growth, enhance societal resilience in the face of technological change and pave the way for future-ready education systems for the next generation of workers. Eight Futures of Work: In a complementary report— —we have imagined various Scenarios and Their Implications scenarios for what the future of work might look like by the year 2030 and what the key implications are for actions today. Unsurprisingly, the need to anticipate changes in the labour market, prepare for reskilling—that is, giving workers the skills and capabilities needed for the future workplace—and support job transitions all emerge as prominent priorities. Yet while there has been much forecasting on transformations in labour markets, few practical approaches exist to identify reskilling and job transition opportunities. This report provides a valuable new tool that will help individual workers, companies, and governments to prioritize their actions and investments. Towards a Reskilling Revolution: A Future of Jobs for All introduces a new approach to identifying reskilling and job transition opportunities, including those that might not be immediately apparent. Using big data analysis of online job postings, the methodology in this report demonstrates the power of a data-driven approach to discover reskilling pathways and job transition opportunities. The methodology can be used to inform the actions of individual workers, policy-makers and companies. It can be applied to a variety of taxonomies of job requirements and sources of data. A Future of Jobs for All 01

6

7 Towards a Reskilling Revolution: A Future of Jobs for All The key question, then, for both individuals and employers Introduction facing these disruptions—and for governments and other The path to a good life appears increasingly difficult to identify stakeholders seeking to support them—is how to better and attain for a growing number of people across our global anticipate and proactively manage the current realignments and community. Gender, inter-regional, generational and income transitions of the labour market to shape a future of work that inequalities are at risk of widening. A key factor driving these expands economic growth and opportunities for all. concerns is the changing nature of work and the extent to , developed by the World Towards a Reskilling Revolution which opportunities for finding stable, meaningful work that Economic Forum in collaboration with The Boston Consulting provides a good income have increasingly become fractured Group and Burning Glass Technologies, aims to provide one key and polarized, favouring those fortunate enough to be living in building block for workers looking to find their place in the future 1 certain geographies and to be holding certain in-demand skills. of work and for business leaders and governments looking to Economic value creation is increasingly based on the use of build more prosperous companies and productive economies ever higher levels of specialized skills and knowledge, creating and societies. Using the labour market of the United States as an unprecedented new opportunities for some while threatening example, the report introduces an innovative, big data approach to leave behind a significant share of the workforce. In a recent built on conventional labour market information systems as well survey of OECD countries, more than one in four adults reported as online job postings. It demonstrates the power of data-driven a mismatch between their current skill sets and the qualifications approaches for finding solutions to job disruptions, including job 2 required to do their jobs. transition pathways and reskilling opportunities that might not be Even among people formerly working good jobs, disruptive immediately apparent. technological and socio-economic forces threaten to swiftly The methodology introduced in this report can be used outdate the shelf life of people’s skillsets and the relevance of to inform the actions of individual workers, policy-makers and what they thought they knew about the path to social mobility companies. Importantly, it is not limited to the geography or data 3 There is a sense that the rise of and rewarding employment. presented here, and can be feasibly adapted to different jobs artificial intelligence, robotics and other digital developments is and skills taxonomies, divergent demand projections and broadly upending the primacy of human expertise in the economy. The to new sources of data about the labour market. Our aim is to individuals who will succeed in the economy of the future will be inspire similar efforts to think about reskilling and job transition those who can complement the work done by mechanical or opportunities among public and private actors globally. It is our 4 algorithmic technologies, and ‘work with the machines’. hope that the report will become a valuable tool to move beyond Employers, too, are feeling the effects of these changes. the current impasse of polarized job prospects, help individuals ManpowerGroup’s 2017 Talent Shortage Survey found that 40% uncover opportunities to build a good life and, above all, inspire of employers reported difficulties in finding skilled talent, while confidence that lifelong learning and reskilling on a society-wide the number of employers filling these gaps by re-training and scale are truly possible. developing people internally has more than doubled since 2015, This report is structured as follows: The next section 5 from just over one in five to more than half. Even so, the rate of introduces our data-driven approach to mapping job transition change is threatening to outpace employers’ positive efforts. The opportunities, providing a brief overview of the methodological found The Future of Jobs, World Economic Forum’s 2016 report, building blocks and core elements of the approach. The following that, by 2020, across all types of occupations, on average, more section explains how the methodology may be used by policy- than a third of the core skills needed to perform most jobs will be makers, corporate strategic workforce planners and others, using 6 made up of skills currently not yet considered crucial to the job. A Future of Jobs for All 03

8 by the labour market disruptions of the Fourth Industrial data for the United States as an example throughout. The third Revolution. For a detailed, more technical description of our section then demonstrates the relevance of the approach to methodology, please refer to the publication’s Appendix A: Data individuals, putting at their disposal a wide range of job transition and Methodology. pathways according to their own priority criteria. The final section concludes the report by briefly discussing the measures needed to support job transitions and reskilling at scale, and suggesting Is the job transition viable? possible extensions of our work. For the interested reader, a A conundrum often cited in the current debate on the future of methodological appendix provides a detailed, more technical work is the contention that “not every displaced coal miner will be discussion of our approach. 9 Rhetoric aside, how might able to become a software engineer”. one actually go about assessing the practical viability of various theoretical job transition options? Mapping Job Transition From a methodological point of view, what is needed in order to do this is an ability to break down jobs into a series Opportunities of relevant, measurable component parts in order to then Calls for stepping up workforce reskilling as a critical component systematically compare them and identify any gaps in knowledge, of preparing labour markets for the Fourth Industrial Revolution skills and experience. If we were able to do this, it would then have become ever more urgent. Until now, however, few practical become possible to calculate the ‘job-readiness’ or ‘job-fit’ of approaches have existed to identify and systematically map out any one individual on the basis of objective criteria. Furthermore, realistic job transition opportunities for workers facing declining we can think of jobs as a collection of tasks that need to be job prospects, answering the question: “what kinds of jobs 10 accomplished within a company. Viable future employees are could affected workers actually reskill to?”. Accordingly, the those equipped to perform those tasks, individuals who possess aim of this report is to provide a valuable new tool that will help the necessary knowledge, skills, and experience. individual workers, companies, and governments to prioritize For the purposes of this publication we have assumed their actions, time and investments. In particular, the data-driven that those who currently hold jobs that require specific skills approach established in this publication can be used to inform and knowledge typically possess the skills and knowledge in policy-makers, corporate strategic workforce planners and 11 question. Once we know the knowledge and skills requirements individuals about possible pathways to meet the anticipated of a job, we can assume that employees transitioning out of that labour demands of the future. It maps out opportunities for job job will be able to bring those capacities into any new roles. transitions for workers currently holding jobs that are highly likely Therefore, the core of our data-driven approach to assessing to be disrupted by structural shifts in the labour market but also the viability of a job transition consists of calculating the similarity provides a method to anyone looking to upskill and improve their between the requirements of two jobs in order to compute wage prospects and job satisfaction. an objective ‘similarity score’ between them. Similarity scores In this publication, we concentrate on job transitions for express the overlap between the activities or tasks that need workers in the United States whose jobs are expected to to be performed in a role as well as between primary indicators 7 disappear due to technological change in the medium-term. of job-fit such as knowledge, skills and abilities, and between To do this, we use a range of data on US employment in 2016 secondary indicators of job-fit such as years of education and from innovative data sources, as detailed below, as well as years of work experience (see Table 1 for an overview of the projections of expected employment change by 2026 from the components of jobs used in the calculation of similarity scores 8 US Bureau of Labor Statistics. It is important to note that we and the report’s Appendix A: Data and Methodology for a do not ourselves predict changes in demand for certain types of 12 comprehensive technical description). jobs in this publication. Rather, we utilize the US Bureau of Labor To make this type of analysis possible in practice, data Statistics’ official forecast of employment in 2026 as an input to from two distinct sources inform our study: the US Bureau of establish our overall approach. However, the methodology used Labor Statistics’ Occupational Information Network (O*NET) in this publication can be readily adapted to other data sets, or to and Burning Glass Technologies. The Occupational Information various scenarios that imagine higher or lower disruptions in the Network (O*NET) database is the primary source of occupational demand for certain types of jobs. information in the United States and contains information on The purpose of the exercise is to uncover, in a systematic required skills, knowledge, abilities, education, training, education viable way, job transition opportunities that are both desirable and and experience to perform a job; it groups individual jobs into from the point of view of those workers affected by labour market clusters of related professions, or ‘job families’, and is continually disruptions. We develop a number of complementary approaches updated by the US government by surveying a broad range from the perspective of both an individual worker seeking of workers. Burning Glass Technologies is a big data labour guidance on high-quality, stable new job opportunities as well market analysis provider that has compiled a unique data set as from the perspective of a policy-maker or corporate planner aggregating insights from more than 50 million online job postings seeking to optimize the collective outcomes for a wider range of in the United States over a two-year period, between 2016 individuals. and 2017, ‘scraping’ data from approximately 40,000 unique This section presents an overview of our data-driven 13 The database developed by Burning Glass online sources. approach to measuring the viability and establishing the Technologies encompasses information on approximately 15,000 desirability of various job transition options for workers affected Towards a Reskilling Revolution 04

9 Table 1: Components of a job Content Aptitudes Experience Work activities are the is the body of facts, principles, Time spent in education is the duration of time Knowledge range of tasks that need to be theories and practices that acts as a foundation for spent gaining knowledge and skills through a formal skills accomplished within a job role route of training are the time spent Years of work experience Skills are used to apply knowledge to complete forming and improving skills to apply a given tasks knowledge through on-the-job practice Cross-functional skills are skills required by a variety of job roles which are transferrable to a are the share of Years of job family experience broad range of job role work experience to date that has been spent within are particular to an industry Specialized skills related professions which exhibit similarities in their or a job role and are not easily transferable (e.g. required skills, knowledge and overall profile skills related to the use, design, maintenance and repair of technology) Abilities are the range of physical and cognitive capabilities that are required to perform a job role Note: Elaboration based on taxonomies by Burning Glass Technologies and Occupational Information Network (O*NET). Table 2: Examples of high, medium and low similarity jobs Starting job Similarity score Target job ‘Job-fit' category Office Clerks, High 0.92 Municipal Clerks General Medium 0.87 First-Line Supervisors of Office and Administrative Support Workers Low Aerospace Engineering and Operations Technicians 0.81 Cooks, Dining Room and Cafeteria Attendants and Bartender Helpers High 0.93 Fast Food Medium 0.86 Butchers and Meat Cutters Locksmiths and Safe Repairers Low 0.82 Electrical Electrical and Electronics Repairers, Powerhouse, Substation and Relay 0.91 High Engineering Medium 0.86 Geothermal Technicians Technicians Low 0.81 First-Line Supervisors of Agricultural Crop and Horticultural Workers Web Developers Computer 0.92 High Programmers Medium 0.86 Computer and Information Systems Managers Low 0.82 Anthropologists Source data: Burning Glass Technologies and US Bureau of Labor Statistics. unique skills across approximately 550 unique skill clusters create a schema (in essence, a matrix) to identify the job-fit 14 between all 958 jobs in our dataset (see (categorized into baseline, specialized, and software skills). Figure 1 on page 6). The combined data set used in our analysis consistently The resulting similarity scores for each pair have a numeric covers 958 unique types of jobs, as classified by the value between 0 and 1. They can be seen as a proxy measure for 15 representing the Occupational Information Network (O*NET), the feasibility of transitioning between the two jobs. Job pairs that have a similarity score of 1 can be said to have a perfect fit, while large majority of the United States workforce, and provides reliable data points on the various components that define job-fit: job pairs with a similarity score of 0 have the most remote and imperfect fit. For example, a computer programmer and a web work activities, skills, knowledge, abilities, years of experience developer have a high job-fit with a similarity score of 0.92, while and education. Following the method established by Burning an office clerk and an aerospace engineering technician have a Glass Technologies, our study aggregates these components 16 of job-fit into an index of similarity, or ‘similarity scores’. We low job-fit with a similarity score of 0.81 (see ). Table 2 scores as scores of at least We describe use these similarity scores as a tool to objectively measure the high similarity similarity between each pair of our 958 unique job types and 0.9 or higher, medium similarity scores as those between 17 low similarity 0.85 and 0.9, and scores as those below 0.85. A Future of Jobs for All 05

10 Figure 1: Job transition matrix between 958 jobs in the United States Arts, Design, Entertainment, Sports, and Media Building and Grounds Cleaning and Maintenance Computer and Mathematical Architecture and Engineering Business and Financial Operations Construction and Extraction Education, Training, and Library Food Preparation and Serving Transportation Sales and Related Healthcare Practitioners and Technical Installation, Maintenance, and Repair Production Personal Care and Service Office and Administrative Life, Physical, and Social Science Community and Social Service Farming, Fishing, and Forestry Protective Service Architecture and Engineering high similarity score n medium similarity score n Arts, Design, Entertainment, Sports, and Media n low simiarity score Building and Grounds Cleaning and Maintenance Business and Financial Operations Community and Social Service Computer and Mathematical Construction and Extraction Education, Training, and Library Farming, Fishing, and Forestry Food Preparation and Serving Healthcare Practitioners and Technical Installation, Maintenance, and Repair Life, Physical, and Social Science Office and Administrative Personal Care and Service Production Protective Service Sales and Related Transportation Source data: Burning Glass Technologies and US Bureau of Labor Statistics. depicts examples of jobs that have high, medium and Table 2 For example, prospective job movers are unlikely to be hired low levels of similarity. It indicates that a job pair is most likely to when their work experience and educational background are have a degree of job-fit that would enable a viable job transition if significantly divergent from the requirements of a job. The US Figure 1 depicts the Bureau of Labor Statistics’ Occupational Information Network similarity scores are at least 0.85 or above. (O*NET) provides a reasonable measure of this profile, in the overall job-fit matrix between all 958 types of jobs (categorized form of so-called ‘job zones’. Job zones capture an occupation’s by job family) in the United States in our dataset. Where a zone expected level of education, related experience, and on-the- is highlighted in dark blue, the corresponding row and column job training required to perform a job. They are measured on a define two occupations with a combined profile that suggests a 1-to-5 scale, where occupations in job zone 1 require little or no high degree of job-fit. preparation (for example dish washers) and occupations in job By themselves, similarity scores provide a useful tool for zone 5 require extensive preparation (for example molecular and a systematic and comprehensive comparison of job-fit and cellular biologists). By restricting job zone changes to no more for identifying viable job transition options. However, as with than -1 or +1, our analysis allows us to control for unrealistic or any composite index, the scores provide a highly aggregated summary view of the theoretical viability of any given job unrewarding moves. The restriction also ensures consistency in the actual level of skills and knowledge use within any given transition. Additional filter criteria are needed to ensure that the job-fit indicated by the aggregate similarity score stays realistic. occupation. Towards a Reskilling Revolution 06

11 To summarize, in order to be able to say that a job transition Finding Job Transition viable job transition option, we opportunity represents a require a pairing of a starting job and target job that involves: Pathways for All (1) a medium or high level of job-fit and (2) realistic leaps in Having established the parameters for viable and desirable job expected years of education or work experience. transition options, we now turn to demonstrating how our data- driven approach may be operationalized to map the opportunity Is the job transition desirable? space for job transitions and create a practical compendium Within the full range of possible job transitions, there are a of job transition options throughout the current—and future— number of transitions that may be viable options—in the sense labour market in the United States. We present two distinct but detailed in the previous section—but which are nevertheless complementary lenses that utilize our principles of viable and unlikely to represent sustainable or attractive options for the desirable job transition options to speak to the concerns and individuals seeking to move jobs concerned. Two parameters priorities of a number of different stakeholder groups across the capture these concerns: the long-term stability of the target job employment ecosystem: and its capacity to financially uphold (or improve) the standard of leadership lens — A that provides policy-makers or corporate living to which the prospective job mover is currently accustomed. planners with a practical tool for maximizing productive Some theoretically viable job transitions are unsustainable re-deployment opportunities for workers affected by labour and undesirable simply because the number of people market disruptions and identifying priority job transition projected to be employed in this job category is set to decline. pathways among a number of viable and desirable options, In the medium term, a number of current occupations in the with a view to optimizing the collective outcomes for a wide United States are forecast to shrink or fully disappear due range of individuals. 18 To identify job transitions that are to technological change. undesirable due to declining target job numbers, we have — An individual lens that maps out viable and desirable job used US employment figures for 2016 as well as projections of transition options from the perspective of a single role and expected employment change by 2026 from the US Bureau of measures the size of the opportunity space for affected 19 Labor Statistics. As mentioned previously, in this publication workers contemplating their personal strategy for moving we defer to the US Bureau of Labor Statistics’ official forecast of out of declining job types and navigating more securely the employment in 2026 as a baseline for our analysis. That is to say, uncertainties of the future of work. we do not ourselves predict changes in demand for certain types Throughout our empirical analysis of the United States labour of jobs in this publication. market, we also highlight a range of thought-provoking examples Another type of theoretically viable job transition that is likely of job transition opportunities uncovered by the analysis that to appear less than attractive to prospective job movers despite might not be immediately apparent. a high job-fit involves target jobs whose remuneration fails to match the standard of living afforded by an individual’s current job. Job transitions in which job movers experience a protracted Leadership Lens fall in wages are unlikely to motivate further reskilling efforts or increases in productivity and job satisfaction by the individuals Intended as a practical planning tool for government and concerned. Wage-losing job transitions also present a less-than- business decision-makers, the leadership lens perspective optimal outcome for government efforts in the field of reskilling, can be used to generate an economy-wide simulation of the as public returns on investment through income or consumption ideal pathway of viable and desirable job transitions that would 20 related taxes will fall with employee wages. maximize the resulting job-fit with target jobs to ensure stable and To summarize, in order to be able to say that a viable job good quality future employment for affected workers currently desirable job transition transition opportunity represents a holding jobs that are set to become obsolete due to structural option, we require a pairing of a starting job and target job that shifts in the labour market. Job transitions are simulated using a 21 involves: (1) stable long-term prospects, i.e. a job transition into linear optimization algorithm. an occupation with job numbers that are forecast not to decline; To operationalize this approach for the United States, we have and (2) wage continuity (or increases), i.e. a level of employee used the official ten-year forecast of employment change produced 22 remuneration for tasks performed in the new job that does not biennially by the Bureau of Labor Statistics. There continues to fall below a level that would allow the individuals concerned to be considerable debate about the degree of disruption to jobs that maintain their current standard of living. is likely to occur across global labour markets in the coming years. Our use of the 2026 Bureau of Labor Statistics data should not necessarily be considered an endorsement of these projections by the World Economic Forum. Indeed, the data-driven approach presented here could plausibly be executed using other forecasts, as long as sufficiently detailed data exists. The Bureau of Labor Statistics projections predict that, over the period up to 2026, the US labour market will see a structural employment decline of 1.4 million redundant jobs, against structural employment growth of 12.4 million new jobs A Future of Jobs for All 07

12 Table 3: Snapshot of projected US job changes by 2026 Employment Change in employment Gender breakdown in 2016 (%) (thousands) 2016–2026 (thousands) Job family Female Male 2016 2026 Increasing jobs Declining jobs Net change Office and Administrative 66 34 22,621 22,730 751 –642 109 436 –41 15,523 15,088 54 477 Sales and Related 46 Business and Financial Operations 49 13,578 14,865 1,334 –48 1,286 51 –33 52 48 13,436 14,688 1,286 1,252 Food Preparation and Serving Healthcare Practitioners and Technical 66 34 12,917 15,246 2,339 –10 2,330 Transportation 16 84 10,266 10,907 650 –9 640 8,926 –511 75 8,558 –368 142 Production 25 789 –4 793 9,317 8,528 38 62 Education, Training and Library Construction and Extraction 3 97 7,157 7,955 800 –1 799 55 45 Personal Care and Service –1 1,164 1,165 7,516 6,352 383 –29 411 6,111 Installation, Maintenance and Repair 5 95 5,729 Building and Grounds Cleaning and 5,619 24 76 6,109 490 0 490 Maintenance 660 4,765 5,402 –23 638 Computer and Mathematical 29 71 196 76 3,573 24 Protective Service 154 –42 3,419 197 Architecture and Engineering 16 84 2,689 2,886 197 0 61 343 –3 346 2,866 2,523 39 Community and Social Service Arts, Design, Entertainment, Sports and Media 60 146 –26 172 2,567 2,421 40 2,045 –14 81 2,113 67 Farming, Fishing and Forestry 20 80 125 1,436 125 0 42 Life, Physical and Social Science 58 1,311 149,389 160,368 12,416 –1,437 10,979 37% 63% Total US Bureau of Labor Statistics. Source data: The figures above exclude 4% of US employment, due to differences in SOC and O*NET job categorization. Note: Figure 2: Projected structural changes in the US job market Table 3 and (see Figure 2 ). According to this forecast, only one by 2026 job family—Production—will experience an overall net job decline. However, both Production and Office and Administrative roles Healthcare Practitioners and Technical Business and Financial Operations are set to experience a significant employment decline. Unlike Food Preparation and Serving Production, however, the Office and Administrative job family is Personal Care and Service Construction and Extraction forecast to experience sufficient new job gains as well in roles Education, Training and Library like Billing, Cost and Rate Clerks, Receptionists and Information Office and Administrative Computer and Mathematical Clerks, and Customer Service Representatives to counter-balance Transportation the shrinking of other occupational categories, such as Data Entry Building and Grounds Cleaning and Maintenance Sales and Related Keyers, File Clerks, Mail Clerks, and Administrative Assistants. Installation, Maintenance and Repair The optimization algorithm used for our analysis maximizes Community and Social Service job-fit between starting and target jobs, and therefore the actual Architecture and Engineering Protective Service feasibility of job transition options across all of the 958 job Arts, Design, Entertainment, Sports and Media types in our data set, representing the large majority of the US Production Life, Physical and Social Science workforce. Job transition options are filtered according to viability Farming, Fishing and Forestry and desirability criteria. Transitions are excluded as unviable if 2,500 2,000 -1000 -500 0 1,500 1,000 500 they would require moving to a target job with a low similarity Job gain/decline (thousands) score or if they would require moving to a target job demanding Burning Glass Technologies and US Bureau of Labor Statistics. Source data: vastly different levels of education and experience. Job transitions Note: The figures above exclude 4% of US employment, due to differences in SOC and O*NET job categorization. are only enacted towards target jobs that would be desirable, with total employment in the target job remaining stable or 08 Towards a Reskilling Revolution

13 Figure 3: Optimized viable and desirable job transitions across job families by 2026 Target job family Starting job family Architecture and Engineering Arts, Design, Entertainment, Sports and Media Building and Grounds Cleaning and Maintenance Business and Financial Operations Community and Social Service Computer and Mathematical Construction and Extraction Education, Training and Library Farming, Fishing and Forestry Food Preparation and Serving Healthcare Practitioners and Technical Installation, Maintenance and Repair Life, Physical and Social Science Office and Administrative Personal Care and Service Production Protective Service Sales and Related Transportation Viable job transition options found Gross job destruction by 2026 Disrupted jobs without viable transition options Architecture and 0.0 0.0 N/A Engineering Arts, Design, Entertainment, 5.2 –26.2 21.0 0.9 0.1 4.5 1.4 1.0 0.1 0.9 0.1 0.1 11.9 Sports, and Media Building and Grounds 0.0 0.0 N/A Cleaning and Maintenance Business and Financial 36.9 –47.8 10.9 36.9 Operations Community and Social 3.0 –3.0 0.0 Service Computer and 22.6 –22.6 0.0 22.6 Mathematical Construction and 0.0 1.2 –1.2 0.1 0.1 0.4 0.2 0.1 0.3 Extraction Education, Training, and 0.0 –3.9 3.9 3.9 Library Farming, Fishing, and 0.4 13.8 –14.2 3.5 0.1 9.2 1.0 Forestry Food Preparation and 33.3 –33.3 0.0 30.2 3.1 Serving Healthcare Practitioners 7.8 –9.8 2.0 1.7 6.1 and Technical Installation, 3.5 25.4 –28.9 0.6 1.4 4.5 0.9 2.9 0.0 1.4 13.7 Maintenance, and Repair Life, Physical, and 0.0 N/A 0.0 Social Science Office and –642.0 20.2 621.8 2.5 8.2 8.8 30.5 0.0 5.0 2 3 6 .1 7. 6 0.4 5.7 40.4 8.0 2 21.1 2.0 11. 8 20.9 13.0 Administrative Personal Care and 0.6 0.0 –0.6 0.2 0.4 Service 20.8 – 510.7 489.9 0.9 1.1 5 .1 2.0 0.4 3.0 2.1 60.9 Production 5.2 6.7 0.6 298.6 2 7.1 20.2 11. 0 13.2 10.5 21.4 41.7 –41.7 0.0 Protective Service 0.7 0.3 2.4 34.8 3.5 –41.3 41.2 0.1 4.7 Sales and Related 0.6 3.2 2.7 0.5 29.5 1.0 –9.4 8.4 0.2 0.2 5.5 1.5 0.4 0.6 Transportation Optimal number 67.2 –1,436.6 1,369.4 256.3 33.5 16.7 19.8 16.5 264.4 2.6 41.0 324.1 14.3 40.5 100.1 24.3 80.6 13.6 45.6 16.4 7.1 52.2 of transitions to job family by 2026 Gross job 660.2 649.7 476.9 142.4 197.2 172.3 489.6 1,333.9 346.1 195.6 799.9 793.3 81.4 1,285.5 2,339.3 411.4 124.8 751.3 1,164.9 12,415.7 creation by 2026 Source data: Burning Glass Technologies and US Bureau of Labor Statistics. Note: Units = 1,000s of people. A Future of Jobs for All 09

14 increasing through 2026 and the difference in wages between an opportunities—are within the Construction and Extraction job 23 individual’s old and new jobs remaining neutral or positive. family and involve target jobs such as Construction Laborers Given the above conditions, the optimization algorithm and Electricians. The next largest opportunity, amounting to about 60,000 well-fitting jobs, is in the Installation, Maintenance used for our analysis is able to find ‘good-fit’ job transitions for and Repair job family, followed by transition options in the vast majority of workers currently holding jobs experiencing Farming, Forestry and Fishing, Transportation, and Office and technological disruption—96%, or nearly 1.4 million individuals. Administrative roles. Once all ‘good-fit’ job transition options highlights suggested ‘good-fit’ job transitions between Figure 3 within the Production job family are taken into account, disrupted and across job families uncovered by our optimization algorithm. workers left without viable or well-fitting new opportunities The light shades indicate situations in which there are only a amount to about 21,000 individuals—or around 4% of the current small number of suggested ‘good-fit’ transition options between workforce in those roles. job families (or none at all) while the dark shades indicate larger Across other job families, growth in demand for Healthcare numbers of transition options within job families. Interestingly, the majority of ‘good-fit’ job transition options—70%—will require the Practitioners and Technicians may absorb some of the structural decline within employment in Food Preparation and Serving job mover to shift into a new, hitherto often unfamiliar cluster of roles, i.e. a new job family. Such job family shifts are the result of Related roles with ‘good-fit’ new opportunities. Technological disruptions within the Computer and Mathematical job family may structural employment decline in particular starting job families, by the availability of ‘better-fit’ target jobs outside the starting be balanced out by transition options within the same job family, while displaced workers in Business and Financial Operations job family, and by the occurrence of employment growth in job may similarly find some ‘good-fit’ new opportunities within families other than the starting one. For example, for roles in the their own job family—but approximately 11,000 of the 48,000 Production job family, such as Electromechanical Equipment displaced workers in Business and Financial Operations workers Assemblers, opportunities can be found in the Architecture and are left with no ‘good-fit’ viable transition options. Engineering job family in positions such as Robotics Technicians In all, our leadership lens optimization model uncovers that and Civil Engineering Technicians. A smaller number—30%—of approximately 4.7% of all US workers projected to be displaced workers holding jobs in structural decline have viable ‘good-fit’ by structural labour market shifts by 2026—approximately 57,000 job transition opportunities within their own current job family. For individuals—are left without immediately viable job transition example, Data Entry Keyers whose jobs are being disrupted by options. Across all job families, the affected workers are heavily technology can transition to becoming Medical Secretaries. Both concentrated in three roles: Postal Service Mail Sorters (Office roles are within the Office and Administrative job family. and Administrative job family), Processing Machine Operators According to the Bureau of Labor Statistics forecast, (Production job family), and Sewing Machine Operators occupations in the Office and Administrative and Production job families will experience the highest rate of job disruption by 2026, (Production job family), precisely the kind of occupations accounting for a combined 1.15 million jobs lost to structural predicted to be heavily impacted by increasing workplace Table 4 labour market change, or 80% of the total (see ). automation. Within our leadership lens optimization model, 238,000 of the It should be noted that the difficulty of finding ‘high-fit’ job 642,000 total current workers within the Office and Administrative transition options depends on the strictness of the initial criteria job family that require new opportunities may find well-fitting used. For example, a slightly modified version of our optimization transition options within their own Office and Administrative job model relaxes the conditions for wage stability and prioritizes family. For those who will need to move to another job family to moving workers into new viable ‘good-fit’ jobs even at the price find a well-fitting job, the largest opportunity lies in the Business of accepting lower wages. Once we relax this criterion on the desirability of target jobs we are able to find opportunities for a and Financial Operations job family, amounting to an additional 221,000 viable job transition options and featuring roles such wider range of workers. In the first model, which optimizes job transition options for viable and desirable conditions including as Human Resource Specialists and Real Estate and Property wages, 4.7% of US workers who will need to change jobs Managers. Smaller clusters of job transition opportunities due to future displacement cannot be placed in ‘good-fit’ new also exist in the Sales and Related, Food Preparation and opportunities. If we optimize for wider labour market inclusion Serving Related and Construction and Extraction job families. and accept that some workers can experience wage loss, that Once all ‘good-fit’ job transition options within the Office and figure falls to 3.7%. Administrative job family are taken into account, disrupted Under the more stringent requirement that our optimization workers left without viable or well-fitting new opportunities pathways should maintain or grow workers’ current level of amount to about 20,000 individuals—or around 3% of the current wages, ‘good-fit’ job transition opportunities are likely, on workforce in those roles. average, to be located in target jobs that require approximately The Production job family is similarly expected to be heavily two years of additional education and two years of additional disrupted by the Fourth Industrial Revolution, with 511,000 jobs work experience. When relaxing the stable wage constraint, expected to be displaced. Unlike in the case of the Office and Administrative job family, however, the Production job family is on average, this experience gap falls to one year of additional ). While there are undoubtedly not expected to create a significant number of viable or desirable Table 5 education required (see intra-job family transition opportunities. The largest opportunities benefits to placing a larger number of individuals in new roles and finding transition opportunities that require more similar levels of for displaced Production workers uncovered by our optimization model—amounting to approximately 299,000 ‘good-fit’ transition education and work experience, our analysis finds that, under a Towards a Reskilling Revolution 10

15 Table 4(a): ‘Good-fit’ job transition options for roles within the Office and Administrative and Production job families Production Target job Starting job 'good-fit' transition opportunities 140,000 Assembly Line Workers Construction Labourers Electricians 45,000 Electrical and Electronic Equipment Assemblers Inspectors, Testers, Sorters, Samplers and Weighers Production, Planning and Expediting Clerks 18,000 Farm and Ranch Managers 17,000 Printing Press Operators First-Line Supervisors of Helpers, Labourers, Inspectors, Testers, Sorters, Samplers and Weighers 15,000 and Material Movers, Hand Molding, Coremaking and Casting Machine Setters, 15,000 Industrial Machinery Mechanics Operators and Tenders, Metal and Plastic Extruding and Drawing Machine Setters, Roustabouts, Oil and Gas 11,000 Operators and Tenders, Metal and Plastic Cutting, Punching and Press Machine Setters, Construction Labourers 11,000 Operators and Tenders, Metal and Plastic Electromechanical Equipment Assemblers Electricians 8,000 Grinding, Lapping, Polishing and Buffing Machine Tool Sheet Metal Workers 8,000 Setters, Operators and Tenders, Metal and Plastic Pipelayers 8,000 Structural Metal Fabricators and Fitters Aircraft Structure, Surfaces, 7,000 Structural Iron and Steel Workers Rigging and Systems Assemblers Inspectors, Testers, Sorters, Samplers and Weighers Quality Control Analysts 7,000 Farm and Ranch Managers 7,000 Prepress Technicians and Workers Civil Engineering Technicians 7,000 Inspectors, Testers, Sorters, Samplers and Weighers Engine and Other Machine Assemblers Electricians 7,000 Cutting, Punching and Press Machine Setters, 6,000 Tile and Marble Setters Operators and Tenders, Metal and Plastic Automotive Body and Related Repairers 6,000 Grinding and Polishing Workers, Hand Tool and Die Makers 5,000 Industrial Machinery Mechanics Photographic Process Workers and Processing Computer User Support Specialists 5,000 Machine Operators Source data: Burning Glass Technologies and US Bureau of Labor Statistics. A Future of Jobs for All 11

16 Table 4(b): ‘Good-fit’ job transition options for roles within the Office and Administrative and Production job families Office and Administrative Target job 'good-fit' transition opportunities Starting job Secretaries and Administrative Assistants, 69,000 Billing, Cost and Rate Clerks Except Legal, Medical and Executive First-Line Supervisors of Office and Secretaries and Administrative Assistants, 51,000 Except Legal, Medical and Executive Administrative Support Workers Executive Secretaries and Human Resources Specialists 39,000 Executive Administrative Assistants Legal Secretaries 34,000 Paralegals and Legal Assistants Executive Secretaries and Property, Real Estate and 31,000 Community Association Managers Executive Administrative Assistants Customer Service Representatives 31,000 Office Clerks, General First-Line Supervisors of Food Preparation Tellers 31,000 and Serving Workers Executive Secretaries and 28,000 Administrative Services Managers Executive Administrative Assistants Medical Secretaries Data Entry Keyers 27,000 24,000 Bookkeeping, Accounting and Auditing Clerks Accountants Real Estate Sales Agents 22,000 Word Processors and Typists Executive Secretaries and 17,000 Training and Development Specialists Executive Administrative Assistants Network and Computer Systems Administrators 12,000 Computer Operators Secretaries and Administrative Assistants, Meeting, Convention and Event Planners 12,000 Except Legal, Medical and Executive Tellers 11,000 Opticians, Dispensing Interviewers, Except Eligibility and Loan Data Entry Keyers 10,000 Postal Service Mail Carriers Brickmasons and Blockmasons 10,000 First-Line Supervisors of Retail Sales Workers 9,000 Office Machine Operators, Except Computer Eligibility Interviewers, Government Programsw 9,000 File Clerks Secretaries and Administrative Assistants, Paralegals and Legal Assistants 8,000 Except Legal, Medical and Executive Source data: Burning Glass Technologies and US Bureau of Labor Statistics. Towards a Reskilling Revolution 12

17 Table 5: Comparison of outcomes with different priority criteria With stable wage requirement With no wage restrictions Outcomes Job transitions with ‘good-fit’ options (millions/share of workers) 1.369 (95.3%) 1.383 (96.3%) Job transitions without ‘good-fit’ options (millions/share of workers) 0.067 (4.7%) 0.053 (3.7%) Share of workers needing to move to new jobs who are female 57% 57% 71% 70% Share of job transitions that involve a change in job family 65% 100% Share of job transitions with stable or increasing wages $19,000 $15,200 For those increasing, average annual wage increase — Share of job transitions with reduction in wages 35% For those decreasing, average annual wage decrease — $8,600 2.0 Average additional years of work experience required 1.7 1.0 2.0 Average additional years of work education required Source data: Burning Glass Technologies and US Bureau of Labor Statistics. wage-agnostic model, wages tend to polarize: 65% of workers considering possible job transition options for at-risk roles, it is experience a sizable average wage increase of about US$19,000, critical to consider the elasticity of opportunity under different Figures 4(a) and 4(b) while a not-insignificant proportion of workers—35%—will need to present a summary overview conditions. of how the number of viable job transition options expands and accept an average pay cut of US$8,600. Conversely, as shown , in an optimization model that does not accept wage Table 5 in contracts in relation to various desirability criteria. Initially, we only consider as a requirement that job demand should not fall cuts, average wages increase by a more modest US$15,200 but exclude the requirement that wages should remain stable for a solid proportion of individuals. Relaxing or restricting other or increase. We then, in turn, tighten different requirements to conditions such as different requirements for work experience find better-fit opportunities, for example imposing two types of and education would also change the results in terms of job placement. wage constraints and a constraint around job fit. The condition Finally, our systemic leadership lens view of job transitions that constrains the number of job transitions the most is that enables us to consider additional dimensions of desirable job workers should look to only move to jobs with high job similarity, transition pathways, such as an integrated lens on gender parity. suggesting that to uncover a larger set of opportunities, reskilling is key. If we look for good-fit jobs with high levels of similarity, Among the workers affected by labour market disruptions, under both models, a larger share—57%—are projected to be female. 16% of roles have no opportunities for transition, and 41% have at In a model allowing wage cuts as well as increases, job transition most three other options. options for displaced women are associated with increasing Of the 1.4 million jobs, which are projected by the US Bureau of Labor Statistics to become disrupted between now wages for 74% of all cases, while the equivalent figure for men is and 2026, the majority – 57% – belong to women. Reflecting only 53%. This trend points to a potential convergence in women and men’s wages among the groups that make job transitions, gender gaps analyzed in the World Economic Forum’s Global Gender Gap Report 2017, partly addressing current wage inequality. the roles that men and women perform in organizations remain out of balance. In today’s US economy, some professions predominantly employ female Individual lens workers, others predominantly male workers. Female workers Intended as a practical guide to uncover the range of job transition opportunities for those threatened by job disruption, Figure 4(a): Average number of job transition options the report’s individual lens perspective aims to highlight viable under different conditions and desirable job transition options from the point of view of individual workers. While the leadership lens presented a model Viable job transition options in which we sought to maximize opportunities for everybody, the individual lens presents the perspective faced by workers in Viable job transition options, any given occupation which is set to experience job losses. To stable or increasing wage do this, we examine job transition opportunities for a number of Viable job transition options, selected jobs across various job families. Taken together, these wage increase of 5% or more examples illustrate the wide range of job transition opportunities for occupations which are set to experience near or medium-term Viable job transition options, high 'job-fit' disruption. 20 30 10 40 50 0 The average worker in the US economy has 48 viable Number of job transitions job transitions, but that figure falls to half that amount if they Burning Glass Technologies and US Bureau of Labor Statistics. Source data: are looking to maintain or increase their current wages. In A Future of Jobs for All 13

18 Figure 4(b): Distribution of job transition options fulfilling stated criteria (all occupations) Viable job transition options Viable job transition options, stable or increasing wage Number of occupations Number of occupations 600 600 500 500 400 400 300 300 200 200 100 100 0 0 150 200 0 50 100 200 150 50 0 100 Number of opportunities Number of opportunities Viable job transition options, high ‘job-fit’ Viable job transition options, wage increase of 5% or more Number of occupations Number of occupations 600 600 500 500 400 400 300 300 200 200 100 100 0 0 0 100 150 200 50 200 100 50 0 150 Number of opportunities Number of opportunities Burning Glass Technologies and US Bureau of Labor Statistics. Source data: options for at-risk roles, it is critical to consider the elasticity of dominate secretarial and administrative assistant roles. In the opportunity under different conditions. US economy 164,000 female workers in those roles are at risk. We know that workers facing job disruption are likely to Some occupations such as assembly line workers predominantly want to or have to move jobs or even change careers. However, employ male workers, and in the United States over 90,000 this method is also intended as a long-term planning tool for workers employed there are at risk. Without reskilling, on individuals that hope to take charge of improving their long-term average, professions that are predominantly female and at risk of career prospects through continuous acquisition of new skills and disruption have only 12 job transition options while at-risk male- relevant experience. As the notion of a job for life increasingly no dominated professions have 22 options. On the other hand, with longer exists, the application of our data-driven approach can reskilling, women have 49 options while predominantly male uncover the opportunities and options available to any individual professions at risk of disruption have 80 options. In other words, for lifelong learning and periodic job transitions. reskilling can narrow the options gap between women and men. Methodologically, our data-driven analysis of individual job More broadly, when considering pathways in an already disrupted transitions between a pair of starting and target jobs can be future of jobs, an opportunity presents itself to close persistent extended, and repeated regularly, to cover a full chain of job gender wage gaps. transition pathways. Job transition pathways illustrate potential Our analysis of opportunities across an individual worker’s long-term reskilling trajectories where a second job transition full profile of available job transition options reveals the distinctive occurs after an initial job transition. Job transition pathways trade-offs which are likely to be experienced by employees allow the discovery of unexpected high-return career trajectories seeking transition opportunities from the vantage point of and reveal that while some job transition options may initially be any given starting job. In considering possible job transition Towards a Reskilling Revolution 14

19 Figure 5(a): Examples of Pathways for Secretaries and Administrative Assistants Insurance Claims Clerks Office and Administrative Occupations 44 wage: $41,000 opportunities similarity score: 0.86 with pay rise Library Assistants, Clerical Office and Administrative Occupations 8 wage: $27,000 Secretaries and opportunities similarity score: 0.89 Administrative with pay cut Assistants Office and Administrative Production, Planning & Expediting Clerks Logisticians Occupations Business and Financial Operations Office and Administrative Occupations wage: $36,000 wage: $49,000 Occupations wage: $78,000 similarity score: 0.91 similarity score: 0.92 Recycling Coordinators Concierges Transportation Occupations Personal Care and Service Occupations wage: $50,000 wage: $31,000 similarity score: 0.89 similarity score: 0.90 Figure 5(b): Examples of Pathways for Cashiers Reservation and Transportation Ticket Agents and Travel Clerks 34 Office and Administrative Occupations opportunities wage: $38,000 with pay rise similarity score: 0.92 Hosts and Hostesses, Restaurant, Lounge and Coffee Shop 4 Food Preparation and Serving Occupations opportunities wage: $21,000 with pay cut similarity score: 0.93 Cashiers Sales and Related wage: $22,000 First-Line Supervisors of Retail Salespersons Sales and Related Occupations Retail Sales Workers wage: $27,000 Sales and Related Occupations similarity score: 0.94 wage: $44,000 similarity score: 0.92 Baristas Food Service Managers Food Preparation and Serving Occupations Food Preparation and Serving Occupations wage: $56,000 wage: $21,000 similarity score: 0.86 similarity score: 0.95 Job Job family Key Remuneration Similarity score with previous job Burning Glass Technologies and US Bureau of Labor Statistics. Source data: A Future of Jobs for All 15

20 Figure 5(c): Examples of Pathways for Bookkeeping, Accounting & Auditing Clerks Brokerage Clerks Office and Administrative Occupations 14 wage: $52,000 opportunities similarity score: 0.88 with pay rise Library Assistants, Clerical Office and Administrative Occupations 6 wage: $27,000 Bookkeeping, opportunities similarity score: 0.85 Accounting and with pay cut Auditing Clerks Office and Administrative Eligibility Interviewers, Title Examiners, Abstractors Occupations and Searchers Government Programs wage: $40,000 Office and Administrative Occupations Business and Financial Operations Occupations wage: $51,000 wage: $44,000 similarity score: 0.88 similarity score: 0.95 Court Clerks Paralegals and Legal Assistants Office and Administrative Occupations Business and Financial Operations wage: $39,000 Occupations similarity score: 0.86 wage: $53,000 similarity score: 0.91 Figure 5(d): Examples of Pathways for Assembly Line Workers Rail Car Repairers Installation, Maintenance and Repair Occupations 59 wage: $54,000 opportunities similarity score: 0.89 with pay rise Packers and Packagers, Hand Transportation Occupations 23 wage: $24,000 opportunities similarity score: 0.89 Assembly Line with pay cut Workers Production Occupations First-Line Supervisors of Construction Construction Labourers wage: $33,000 Construction and Extraction Occupations Trades and Extraction Workers wage: $38,000 Construction and Extraction Occupations similarity score: 0.88 wage: $68,000 similarity score: 0.87 Animal Breeders Nursery Workers Farming, Fishing and Forestry Occupations Farming, Fishing and Forestry Occupations wage: $42,000 wage: $24,000 similarity score: 0.87 similarity score: 0.87 Job Job Family Key Remuneration Similarity score with previous job Source data: Burning Glass Technologies and US Bureau of Labor Statistics. Making Reskilling Real 16 Towards a Reskilling Revolution 16

21 associated with pay cuts, those initial job transition decisions Given the impending job displacement and rapid changes might pave the way to rewarding careers later on. Facing a already underway in the types of skills demanded by the labour markets of the Fourth Industrial Revolution, the arguments for variable horizon of opportunities, aiming for long-term gains after short-term displacement becomes one additional route for taking action now are compelling for individuals, employers and policy-makers: workers with few desirable short term job transition options. Figures 5(a) to 5(d) illustrate selected job transition — particularly those under risk of displacement, individuals, For pathways for a range of jobs at risk from technological disruption. simply to remain employed will require engaging in lifelong For each job, we have defined four profiles (or ‘archetypes’), to learning and regular reskilling. Additionally, for all workers, reflect the range of opportunities—as well as the attitudes and continuous learning will not only be key to securing mindsets with which individuals are likely to approach career employment but also to building stable, fulfilling careers and planning and the lifelong learning and reskilling challenge in the seizing rewarding job transition opportunities. Fourth Industrial Revolution. A first archetype consists of a simple single transition with — For employers, relying solely on new workers entering the a rising wage. A second archetype consists of a single transition labour market with the right ready-made skills will no longer with a declining wage. A third consists of a steady rise in two be sufficient. And while predicting the exact nature of the steps. A fourth consists of an initial decline in the first step demand for skills is impossible, recent research from the followed by an increase. World Economic Forum reveals that across a wide range Secretaries and Administrative Assistants is an occupation of scenarios, investment in workforce reskilling and human for which the United States will see a fall in demand amounting capital development is a ‘no-regret action’—that is, it will be a 24 to 165,000 workers by 2026 according to the Bureau of Labor beneficial investment even in the absence of skills shortages. Statistics. The range of opportunities available to those displaced For — policy-makers, fostering continuous reskilling and workers are illustrated in Figure 5(a) . Despite the magnitude lifelong learning across the economy will be critical in order to of projected losses, Secretaries and Administrative Assistants maintain a labour force with the tools needed to fuel inclusive have 44 viable job transition opportunities which will see them economic growth and to ensure that companies can find retain their current wage or gain in wages, opportunities such workers with the skills needed to help them succeed and as roles as Insurance Claims Clerks or Production, Planning and 25 contribute their full potential to the economy and society. Expediting Clerks. In the long term, those transitions can serve as stepping stones to even more lucrative opportunities, such as In assessing reskilling pathways and job transition roles in Logistics. opportunities with detail and scale, we aim to move the debate Secretaries and Administrative Assistants have a variety of on the future of work to new—and practical—territory. This report opportunities, so it is unlikely that individuals working in those is a beginning. In subsequent publications, the methodology will roles will need to revert to lower paying roles such as Clerical be extended to include additional perspectives and geographies Library Assistants, however, other declining professions have and applied in collaboration with government and business a more constrained horizon. Bookkeeping, Accounting and stakeholders to support workers. Auditing Clerks have only 14 viable opportunities with stable The report points to a number of directions for the efforts or rising wages. A wide range of factors such as the uneven that will be needed to support and scale job transitions and distribution of opportunities geographically mean that workers reskilling efforts, in those roles might need to consider roles with decreasing wages. This will expand their opportunities with good job-fit Planning, delivering and financing reskilling and job to 20. Taking on a role with lesser salary might mean they transitions become Clerical Library Assistants or Court Clerks. Yet those The main limiting factor on opening up a world of job transition roles can serve as a stepping stone to roles that exceed their opportunities is the willingness to make a reasonable investment initial wages—such as Paralegal and Legal Assistants. in reskilling that will bridge workers onto new jobs. While the need for equipping the world’s workforce with the skills for the future of work and emerging job types is clear, the question is what Conclusion policies and strategies may be used to drive and deliver lifelong Current discussions of the future of work have often learning and reskilling at scale. As individuals may need to take emphasized the urgency of reskilling and life-long learning. temporary time out from work while re-training and exploring job Yet, few approaches exist to help identify productive ways of transition options, public as well as private financial support will planning job transitions that can minimize strain on companies’ be needed. Translating reskilling into viable and desirable jobs will workforce strategies, public finances and social safety nets, as require new thinking around workforce planning. As redeploying well as the affected individuals themselves. The purpose of this workers across jobs will become the norm, there will also be report has been to introduce such an approach to mapping a need for agile social protection and insurance mechanisms out job transition pathways and reskilling opportunities, that avoid destabilizing income while prioritizing rapid workforce using the power of digital data to help guide workers, re-integration. Wide-spread adoption of micro-credentials and companies, and governments to prioritize their actions, time new methods of education and training delivery that combine and investments on focusing reskilling efforts efficiently and online and offline models will be necessary for creating new effectively. A Future of Jobs for All 17

22 opportunities for workers. As detailed in two recent World — Geographic expansion: The report’s methodology can be Economic Forum White Papers, extended both to additional geographies outside the United Accelerating Workforce Reskilling for the Fourth Industrial Revolution and Realizing Human States and to cover local geographies—such as the state- level perspective—to help address the needs of local markets , countries such as Potential in the Fourth Industrial Revolution and consider the impact of mobility within and between Denmark and have already seen success experimenting with policy measures that may support the scale of the efforts that geographies when workers move to new jobs. 26 By helping to quantify the gains in aggregate will be required. — The quantification of reskilling efforts: The methodology income of an economy that will result from redeploying workers can be used to assess the amount of time required to make to emerging positions that otherwise might have gone unfilled, job transitions, based on the difficulty of acquiring new skills. the data-driven approach described in this publication is helpful It can also assess the costs associated with reskilling, such as in highlighting the viability of this new vision and in building the actual cost of training and associated opportunity costs to the economic and business case for planning, delivering and determine motivations and incentives. financing reskilling and job transitions. — The Nuanced evaluation of economic benefits: methodology can be used to assess the gains in aggregate Individuals’ mindset and efforts will be key income of an economy that result from job transitions into To even begin thinking about large-scale job transition planning emerging roles that otherwise would have gone unfilled as well and economy-wide reskilling, the role of individuals will be as determine the cost-benefit analysis around government absolutely critical. Some reskilling will require time off work, payments and safety nets (e.g. unemployment benefits). some will require gaining additional formal qualifications, perhaps after decades out of the classroom. These efforts will not be The Different scenarios of changing demand for jobs: — easy, and individuals will need to be adequately supported and methodology can be used to create job transition models as incentivized and will need to be able to see the eventual benefits they apply in different scenarios of growth/decline in jobs (e.g., of continuous reskilling in the form of rewarding job transition a job transition model that proposes that a larger number of pathways. Here, too, the data-driven approach advocated in jobs will be lost as a result of automation). this publication may help to created greater transparency and — The methodology Gender perspectives on job transitions: choice for workers. Nevertheless, what will be required is nothing can be used to promote gender-inclusive proactive workforce less than a societal mindset shift for people to become creative, planning, by uncovering job transition models that promote curious, agile lifelong learners, comfortable with continuous gender equality (as relevant to corporate and policy change. decision-makers). It is our hope that will become Towards a Reskilling Revolution No single actor can solve the job transition and reskilling a valuable tool to move beyond the current impasse of polarized puzzle alone job prospects, help individuals uncover opportunities to build a To make reskilling real, and prepare for accelerated structural good life and, above all, inspire confidence that taking a focused, change of the labour market, a wide range of stakeholders— proactive approach to large-scale reskilling and lifelong learning governments, employers, individuals, educational institutions is truly possible. We also hope it inspires similar efforts to think and labour unions, among others—will need to learn to come practically yet holistically about managing reskilling, upskilling and together, collaborate and pool their resources more than ever job transitions. before. For businesses, working together across traditional industry boundaries and, sometimes, with their competitors, in order to ensure they have the talent for tomorrow they need, Endnotes will hold significant benefits but require new ways of thinking Autor, David, 1 The Polarization of Job Opportunities in the U.S. Labor Market: 27 Governments too will need more rapid and working as well. Center for American Progress and Implications for Employment and Earnings, learning from each other and consider a range of experiments The Hamilton Projects, 2010. for discovering the most effective approaches. Education and 2 McGowan, Müge Adalet and Dan Andrews, Skill Mismatch and Public Policy in OECD Countries , (OECD Economics Department Working Paper No. 1210), training businesses and non-profits will find they are in high OECD, 2015. demand and will need to collaborate with each other—and with 3 Bessen, James, “Toil and Technology: Innovative technology is displacing other stakeholders to determine how they can be most effective. workers to new jobs rather than replacing them entirely”, IMF Finance and Development Magazine , March 2015. Extending the data-driven approach , The Growing Importance of Social Skills in the Labor Market Deming, David, 4 NBER Working Paper 21473, National Bureau of Economic Research, 2015. Data-driven approaches can bring speed and additional value 2016-2017 Talent Shortage Survey , 2 0 17. 5 ManpowerGroup, to reskilling and job transitions. The World Economic Forum will 6 World Economic Forum, The Future of Jobs: Employment, Skills and undertake some of this work in subsequent publications—and we Workforce Strategy for the Fourth Industrial Revolution , 2016. actively encourage others to follow suit. A non-exhaustive list of 7 There are a range of such predictions. See, for example, Frey, Carl Benedikt extensions could look to: and Michael A. Osborne, The future of employment: How susceptible are jobs to computerisation?, Oxford Martin School, University of Oxford, September 2013. However, for the purpose of this study, we use official projections from the US Bureau of Labor Statistics. The proposed model, however, can be applied to any range of predictions with appropriate data. Towards a Reskilling Revolution 18

23 See US Bureau of Labor Statistics, 23 For a detailed, more technical overview of our methodology, please refer to 8 2016–26 Employment Projections and , https://www.bls.gov/emp/ (released on 24 the report’s Appendix A: Data and Methodology. Occupational Outlook Handbook October 2017). Eight Futures of Work: Scenarios and Their 24 World Economic Forum, Implications 9 See, for example, Smiley, Lauren, “Can you Teach a Coal Miner to Code?”, , 2018. Also see: Janah, Leila, “Labor looks different in the 21st Quartz WIRED , 18 November 2015, https://www.wired.com/2015/11/can-you-teach- century—so should job training”, , 2015, http://qz.com/496297/us-job- a-coal-miner-to-code; and Field, Anne, “Turning Coal Miners Into Coders— training-must-catch-up-to-the-work-the-labor-force-will-do; ManpowerGroup, , 2010, http:// , 30 January 2017, https://www. Forbes And Preventing A Brain Drain”, Teachable Fit: A New Approach to Easing the Talent Mismatch www.manpowergroup.com/sustainability/teachable-fit-inside.html. forbes.com/sites/annefield/2017/01/30/turning-coal-miners-into-coders-and- preventing-a-brain-drain. 25 Voss, Eckhard, “Organising Transitions in Response to Restructuring: Autor, David, The “Task Approach” to Labor Markets: An Overview, (NBER Study on instruments and schemes of job and professional transition and 10 Final Report Working Paper No. 18711), National Bureau of Economic Research, 2013, re-conversion at national, sectoral or regional level in the EU”, , http://www.nber.org/papers/w18711. European Commission, 2009. Accelerating Workforce Reskilling for the Fourth A second, perhaps less obvious, assumption is that skill demands for a 11 26 World Economic Forum, Realizing Human Industrial Revolution particular type of job are the same across different firms; see: Deming, David, , 2017; World Economic Forum, Potential in the Fourth Industrial Revolution: An Agenda for Leaders to Shape and Lisa B. Kahn “Skill Requirements across Firms and Labor Markets: Journal of Labor Economics the Future of Education, Gender and Work Evidence from Job Postings for Professionals”, , 2 0 17. , vol. 36, no. S1, 2018, pp. S337-S369. 27 Broadbent, Stefana, “Collective Intelligence: Questioning Individualised Approaches to Skills Development”, in: IPPR and J.P. Morgan Chase New 12 Theoretically, task requirements, knowledge, skills and abilities are sufficient Technology, Globalisation and the Future of Work in elements to uniquely define ‘job-fit’ between any two jobs, however, as these Skills at Work Initiative, , March 2015. Europe: Essays on Employment in a Digitised Economy components are hard to measure with high levels of precision in practice, our calculation of ‘similarity scores’ also includes three secondary dimensions that are more commonly used signals or proxies: time spent in education; years of work experience; and years of experience within the concerned job family. References For a general overview of the innovative new type of big data labour market 13 analysis employed by Burning Glass Technologies and other firms, the reader and Further Reading may refer to: Carnevale, Anthony, Tamara Jayasundera and Dmitri Repnikov, Understanding Online Job Ads Data: A Technical Report , Georgetown Accenture, Burning Glass Technologies and Harvard Business School, Bridge the University Center on Education and the Workforce, April 2014; Reamer, Gap: Re-Building America’s Middle Skills , 2015. Using Real-Time Labor Market Information on A Nationwide Scale: Andrew, “ Acemoglu, Daron and Pascual Restrepo, Robots and Jobs: Evidence from US Exploring the Research Tool’s Potential Value to Federal Agencies and (NBER paper No. w23285), National Bureau of Economic Labor Markets , Credentials that Work Programme, Jobs for National Trade Associations Research, 2017. the Future, April 2013; Manca, Fabio, “Measuring skills shortages in real ti m e”, OECD Skills and Work Blog , 16 March 2016, https://oecdskillsandwork. Acemoglu, Daron and Robert Shimer, “Productivity gains from unemployment wordpress.com/2016/03/16/measuring-skills-shortages-in-real-time/; and insurance”, , vol. 44, 2000, pp. 1195–1224. European Economic Review Wright, Joshua, “Making a Key Distinction: Real-Time LMI & Traditional Labor , 7 February 2012, http://www.economicmodeling. Emsi Blog Market Data”, Autor, David, “Why Are There Still So Many Jobs? The History and Future of com/2012/02/07/making-a-key-distinction-real-time-lmi-traditional-labor- Workplace Automation”, , 2015, http:// Journal of Economic Perspectives market-data. dx.doi.org/10.1257/jep.29.3.3. 14 See Appendix A: Data and Methodology for a detailed description of the , (NBER Working Paper The “Task Approach” to Labor Markets: An Overview ——— , report’s methodology, data and data sources. No. 18711), National Bureau of Economic Research, 2013, http://www.nber. org/papers/w18711. Job types are categorized in accordance with O*NET-SOC 2010 codes for 15 which sufficient information was available from both Occupational Information ——— , The Polarization of Job Opportunities in the U.S. Labor Market: Implications Network (O*NET) and Burning Glass Technologies; https://www.onetcenter. for Employment and Earnings , Center for American Progress and The org/taxonomy.html. Hamilton Project, 2010. While both the underlying components and exact method of calculating 16 Autor, David, et al., “The Skill Content of Recent Technological Change: An similarity scores for this publication are the result of discussions between The Quarterly Journal of Economics Empirical Exploration”, , vol. 18, no. 4, Burning Glass Technologies, World Economic Forum and Boston Consulting 2003, pp. 1279-1333. Group, the approach taken is adaptable and generalizable to other datasets Bessen, James, “Toil and Technology: Innovative technology is displacing and occupational classifications, and may be updated if additional new or workers to new jobs rather than replacing them entirely”, IMF Finance and improved data on jobs becomes available. , March 2015. Development Magazine 17 See the publication’s Appendix A: Data and Methodology for a detailed Bhalla, Vikram, Suzanne Dyrchs and Rainer Strack, “Twelve Forces That Will description of the calculation and categorization of similarity scores. , The Boston BCG perspectives Radically Change How Organizations Work”, 18 The future of employment: Frey, Carl Benedikt and Michael A. Osborne, Consulting Group 2017, https://www.bcg.com/de-de/publications/2017/ , Oxford Martin School, How susceptible are jobs to computerisation? people-organization-strategy-twelve-forces-radically-change-organizations- University of Oxford, September 2013, http://www.oxfordmartin.ox.ac.uk/ work.aspx. downloads/academic/The_Future_of_Employment.pdf. Broadbent, Stefana, “Collective Intelligence: Questioning Individualised Approaches See US Bureau of Labor Statistics, 2016–26 Employment Projections and 19 to Skills Development”, in: IPPR and J.P. Morgan Chase New Skills at Work , https://www.bls.gov/emp/ (released 24 Occupational Outlook Handbook Technology, Globalisation and the Future of Work in Europe: Essays Initiative, October 2017); the US Bureau of Labor Statistics data employed in this , March 2015. on Employment in a Digitised Economy publication predict a structural gross decline in employment of about 1.4 Artificial intelligence and the Brynjolfsson, Erik, Daniel Rock and Chad Syverson. million jobs across a range of occupations until 2026. , (NBER modern productivity paradox: A clash of expectations and statistics Acemoglu, Daron and Robert Shimer, “Productivity gains from unemployment 20 Working Paper No. w24001), National Bureau of Economic Research, 2017. , vol. 44, 2000, pp. 1195–1224. insurance”, European Economic Review Understanding Carnevale, Anthony, Tamara Jayasundera and Dmitri Repnikov, 21 For a detailed, more technical overview of our methodology, please refer to Online Job Ads Data: A Technical Report , Georgetown University Center on the report’s Appendix A: Data and Methodology. Education and the Workforce, April 2014. 22 See US Bureau of Labor Statistics, 2017. Deming, David, The Growing Importance of Social Skills in the Labor Market , (NBER Working Paper 21473), National Bureau of Economic Research, 2015. Deming, David and Lisa B. Kahn, “Skill Requirements across Firms and Labor Markets: Evidence from Job Postings for Professionals,” Journal of Labor Economics , vol. 36, no. S1, 2018, pp. S337-S369. A Future of Jobs for All 19

24 The Future of the Professions: How European Centre for the Development of Vocational Training (CEDEFOP), Susskind, Richard and Daniel Susskind, Technology Will Transform the Work of Human Experts , Oxford University Quantifying Skills Needs in Europe—Occupational Skills Profiles: Press, 2016. Methodology and Application , (CEDEFOP Research Paper No. 30), 2013. European Commission, “EU employment in a global context: where will new Working: People Talk About What They Do All Day and How They Terkel, Studs, , Pantheon Books, 1974. Feel About What They Do jobs come from and what will they look like?” in: Employment and Social Developments in Europe 2013 , 2014, http://ec.europa.eu/social/main.jsp?catI US Bureau of Labor Statistics, Employment Projections 2016-2026 , 2 0 17. d=738&langId=en&pubId=7684&visible=1. Voss, Eckhard, “Organising Transitions in Response to Restructuring: Study Don’t let perfect be the enemy of good: To leverage the data Fengler, Wolfgang, on instruments and schemes of job and professional transition and , Brookings, April 2016, http://www. revolution we must accept imperfection Final Report re-conversion at national, sectoral or regional level in the EU”, , brookings.edu/blogs/future-development/posts/2016/04/14-big-data- European Commission, 2009. revolution-technologies-fengler. Eight Futures of Work: Scenarios and Their Implications World Economic Forum, , Field, Anne, “Turning Coal Miners Into Coders—And Preventing A Brain Drain”, 2018. , 30 January 2017, https://www.forbes.com/sites/annefield/2017/01/30/ Forbes turning-coal-miners-into-coders-and-preventing-a-brain-drain. Accelerating Workforce Reskilling for the Fourth Industrial Revolution: An ——— , , Agenda for Leaders to Shape the Future of Education, Gender and Work The Future of Employment: How Frey, Carl Benedikt and Michael A. Osborne, 2 0 17. Susceptible are Jobs to Computerisation? , Oxford Martin School Programme on the Impacts of Future Technology, September 2013, http://www. The Future of Jobs and Skills in Africa: Preparing the Region for the Fourth ——— , oxfordmartin.ox.ac.uk/downloads/academic/The_Future_of_Employment.pdf. , 2 0 17. Industrial Revolution Goodhue, Dale L. and Ronald L. Thompson, “Task-technology fit and individual The Future of Jobs and Skills in the Middle East and North Africa: Preparing ——— , MIS Quarterly performance”, , vol. 19, no. 2, 1995, pp. 213-236. the Region for the Fourth Industrial Revolution , 2 0 17. Hasan, Sharique, John-Paul Ferguson and Rembrand Koning, The Lives and Realizing Human Potential in the Fourth Industrial Revolution: An Agenda for ——— , , Stanford University, Deaths of Jobs: A Mid-Range Theory of Job Structures , 2 0 17. Leaders to Shape the Future of Education, Gender and Work 2013, https://web.stanford.edu/~sharique/papers/hfk_jobs.pdf. ——— , The Future of Jobs: Employment, Skills and Workforce Strategy for the Hershbein, Brad and Lisa B. Kahn. “Do Recessions Accelerate Routine-Biased Fourth Industrial Revolution , 2016. American Technological Change? Evidence from Vacancy Postings,” , forthcoming NBER Working Paper No. 22762, 2018. The Human Capital Report , 2016. Economic Review ——— , Guidance Note: Anticipating and matching International Labour Organisation (ILO), Wright, Joshua, “Making a Key Distinction: Real-Time LMI & Traditional Labor , 7 February 2012, http://www.economicmodeling. , November 2015, http://www.ilo.org/skills/areas/skills-training- Market Data”, Emsi Blog skills and jobs com/2012/02/07/making-a-key-distinction-real-time-lmi-traditional-labor- for-poverty-reduction/WCMS_534307/lang--en/index.htm. market-data. Janah, Leila, “Labor looks different in the 21st century—so should job training”, , 2015, http://qz.com/496297/us-job-training-must-catch-up-to-the- Quartz work-the-labor-force-will-do. Jeon, Shinyoung, “Enhancing Employment for Women, Youth and Older Workers: Global Talent Competitiveness Index 2014 Why Skills Strategies Matter”, in: , edited by Bruno Lanvin and Paul Evans, INSEAD, 2014, http://global-indices. insead.edu/documents/INSEADGTCIreport2014.pdf. Lorenz, Markus, Michael Rüßmann, Rainer Strack, Knud Lasse Lueth and Moritz Bolle, “Man and Machine in Industry 4.0: How Will Technology Transform BCG perspectives , The Boston the Industrial Workforce Through 2025?”, Consulting Group, 2015, https://www.bcgperspectives.com/content/articles/ technology-business-transformation-engineered-products-infrastructure- man-machine-industry-4/. OECD Skills and Work Manca, Fabio, “Measuring skills shortages in real time”, Blog , 16 March 2016, https://oecdskillsandwork.wordpress.com/2016/03/16/ measuring-skills-shortages-in-real-time/. ManpowerGroup, 2016-2017 Talent Shortage Survey , 2017, http://www. manpowergroup.com/talent-shortage-2016. ——— , Teachable Fit: A New Approach to Easing the Talent Mismatch , 2010, http:// www.manpowergroup.com/sustainability/teachable-fit-inside.html. McGowan, M üge Adalet and Dan Andrews, “Skill Mismatch and Public Policy in OECD Countries”, (OECD Economics Department Working Paper No. 1210), OECD, 2015. Digitalization and the Muro, Mark, Sifan Liu, Jacob Whiton and Siddharth Kulkarni, American workforce , Brookings, 2017. OECD, Skills Matter: Further Results from the Survey of Adult Skills , OECD Skills Studies, 2016, http://dx.doi.org/10.1787/9789264258051-en. Reamer, Andrew, “ Using Real-Time Labor Market Information on A Nationwide Scale: Exploring the Research Tool’s Potential Value to Federal Agencies and , Credentials that Work Programme, Jobs for the National Trade Associations Future, April 2013. Reimsbach-Kounatze, Christian, The Proliferation of “Big Data” and Implications for Official Statistics and Statistical Agencies: A Preliminary Analysis , (OECD Digital Economy Papers, No. 245), OECD, 2015. Schwab, Klaus, The Fourth Industrial Revolution , World Economic Forum, 2016. WIRED Smiley, Lauren, “Can you Teach a Coal Miner to Code?”, , 18 November 2015, https://www.wired.com/2015/11/can-you-teach-a-coal-miner-to-code. Towards a Reskilling Revolution 20

25 Appendix A: Data and Methodology The analysis that forms the basis of this report is based on the Figure A1: Conditions of viable job transitions concept of ‘viable job transitions’, which is comprised of four criteria and explained in more detail below. The concept is Viable job transitions created from a variety of source data. In addition to establishing the overall viability of job transitions, we conduct further specific Condition Main source data analysis on various sub-components of this data. 1. BGT, O*NET Similarity scores between jobs are The majority of our analysis has been conducted using data sufficiently high and from three distinct data sources, as referenced in Figure A1 BGT, O*NET Transition does not require huge 2. explained in detail below. All of the analysis has been conducted leaps in education and experience on data from the United States. Each source provides a different type of job data, allowing us to create an overall combined data set and refine our analysis. Desirable job transitions Data Sources Condition Main source data Occupational Information Network (O*NET) BLS, O*NET 3. Transition involves moving to jobs where numbers are forecast not The Occupational Information Network (O*NET) database is the to decline primary source of occupational information in the United States, Transition leads to a level of wage 4. BGT developed under the sponsorship of the US Department of continuity that allows individuals to Labor/Employment and Training Administration. The database maintain their standard of living groups individual jobs into clusters of related professions, or ‘job families’, and is continually updated by surveying a broad range of workers from each job. Its use in our work is providing both a standardized list of almost one thousand job types, covering the entire US economy, and job-specific descriptors (e.g. required skills and knowledge) on these jobs. Burning Glass Technologies (BGT) The BGT analysis of each job posting results in an The data set compiled by Burning Glass Technologies (BGT) accumulation of detailed information on required skills in each for this report is based on online job postings. This information job. This information is categorized into approximately 15,000 is sourced by ‘scraping’ detailed data for a job from various individual skills within approximately 550 skill clusters (categorized online sources (e.g. job boards, employer sites). The data set into baseline, specialized, and software skills). Information is encompasses detailed information on 958 jobs within the also captured on the education and experience required for a United States. Jobs in the data set are based on standardized job as well as average wages. Additionally, the BGT data set job codes and job titles from O*NET. The data set is based on includes supplementary information on the employment gender approximately 50 million job postings over a two-year period distribution of each job covered from the American Community from 2016 to 2017, covering approximately 40,000 unique data Survey (ACS). sources in the United States. A Future of Jobs for All 21

26 US Bureau of Labor Statistics (BLS) Viable and Desirable Job Transitions: The 2016–2026 National Industry-Occupation Employment Matrix Methodology is developed by the US Bureau of Labor Statistics in the course Condition 1: Similarity scores between jobs of its ongoing Employment Projections program. The 2016 matrix are sufficiently high was developed primarily from the Occupational Employment Assessing viable job transition opportunities requires an Statistics (OES) survey, the Current Employment Statistics (CES) understanding of the requirements necessary to perform a survey, and the Current Population Survey (CPS). The 2016–26 given job and an ability to compare these requirements to the National Employment Matrix encompasses data for approximately requirements of another job. The requirements of a job fall into a 800 jobs in the US and contains information on employment in number of categories: 2016, as well as projections for expected employment in 2026 on an individual job basis. Work activities: — The range of tasks that need to be The information on jobs in the 2016–2026 National accomplished within a job role. Industry-Occupation Employment Matrix is based on Standard Knowledge: — Knowledge is the body of facts, principles, Occupational Classification (SOC) codes. The data set of 958 theories, and practices that acts as a foundation for skills. jobs used in this study captures about 96% of total employment in the 2016–2026 National Industry-Occupation Employment Skills are used to apply knowledge to complete tasks. Skills: — Matrix. Projections of employment per job were developed in » Cross-functional skills: Common, non-specialized skills a series of six interrelated steps, each based on a different required by job applicants to be considered for a role procedure or model and related assumptions: labour force, (applicable to broad categories of jobs). aggregate economy, final demand (GDP) by consuming sector and product, industry output, employment by industry, and » Skills particular to an industry or a job Specialized skills: employment by occupation. The results produced by each that are not easily transferable. For the purpose of refining step are key inputs to following steps, and the sequence may the requirements of a job in the calculations used in this be repeated multiple times to allow feedback and to ensure report, we separate out software skills (the use, design, consistency. maintenance and repair of different types of software). Abilities: — The range of physical and cognitive capabilities that are required to perform a job role. Education: Education is a formal mechanism for acquiring — skills and knowledge. Work and Job Family Experience: Experience plays a — crucial role in forming and improving skills to apply a given knowledge. Table A1: Examples of calibration of similarity scores for high, medium and low similarity jobs Job-fit' category Starting job Similarity score Target job 0.92 Municipal Clerks High Office Clerks, General Medium 0.87 First-Line Supervisors of Office and Administrative Support Workers 0.81 Aerospace Engineering and Operations Technicians Low Dining Room and Cafeteria Attendants and Bartender Helpers High 0.93 Cooks, Fast Food Medium Butchers and Meat Cutters 0.86 Low 0.82 Locksmiths and Safe Repairers Electrical Electrical and Electronics Repairers, Powerhouse, Substation and Relay 0.91 High Engineering Geothermal Technicians Medium 0.86 Technicians 0.81 Low First-Line Supervisors of Agricultural Crop and Horticultural Workers Web Developers High 0.92 Computer Programmers Medium 0.86 Computer and Information Systems Managers Low 0.82 Anthropologists Source data: Burning Glass Technologies and US Bureau of Labor Statistics. Towards a Reskilling Revolution 22

27 Figure A2: Frequency of similarity scores (selected examples) Office Clerks, General Cooks, Fast Food Frequency Frequency 250 350 Aerospace Engineering Locksmiths and Safe Repairers and Operations Technicians 300 0.82 0.81 200 250 First-Line Supervisors of 150 Office and Administrative Butchers and Meat Cutters 200 Support Workers 0.86 0.87 150 Dining Room and 100 Cafeteria Attendants Municipal Clerks and Bartender Helpers 100 0.92 0.93 50 50 0 0 1.00 0.95 1.00 0.75 0.00 0.60 0.65 0.70 0.80 0.00 0.60 0.65 0.70 0.75 0.80 0.85 0.90 0.95 0.85 0.90 Similarity score range Similarity score range Computer Programmers Electrical Engineering Technicians Frequency Frequency 300 350 First-Line Supervisors of Agricultural Crop and Horticultural Workers Anthropologists 300 0.81 0.82 250 250 200 Computer and Information Geothermal Technicians Systems Managers 200 0.86 0.86 150 150 Electrical and Electronics Repairers, Powerhouse, 100 Web Developers Substation and Relay 100 0.92 0.91 50 50 0 0 0.65 0.60 1.00 0.00 0.95 0.85 0.80 0.75 0.90 0.70 0.00 0.65 0.70 0.75 0.80 0.85 0.90 0.95 1.00 0.60 Similarity score range Similarity score range Medium similarity score n High similarity score n n Low similarity score Source data: Burning Glass Technologies and US Bureau of Labor Statistics. Similarity score ranges below 0.60 are excluded, given they are not significant values (frequency of 1). Note: To assess the similarity between the requirements of two Table A1 process across a wide range of examples (see and Figure A2 ). jobs, this report introduces the concept of ‘similarity scores’. Similarity scores express the overlap between requirements such For the purpose of identifying viable job transitions options, as education, experience, training, skills and knowledge, as a we exclude job transitions that are characterized by low similarity scores (below 0.85). numeric value between 0 and 1. They can be seen as a proxy for To arrive at the concept of a numerical similarity score, the feasibility of transitioning between two jobs (or ‘job pair’). Burning Glass Technologies contributes a distinctive approach Job pairs that have a similarity score of 1 share the exact to calculating these scores. This methodology combines data same requirements, while job pairs with a similarity score of 0 from both BGT job posting results and from O*NET’s database of have no requirements in common. For example, a computer job-specific descriptors. In a first step, for each of these two data programmer and a web developer have a similarity score of 0.92, while a computer programmer and an anthropologist only have a sources, individual similarity scores are calculated (‘Burning Glass similarity score of 0.82. For ease of analysis, we have categorized Technologies similarity scores’ and ‘O*NET similarity scores’). similarity scores into high (scores of at least 0.9), medium This is necessary to harness the advantages of both standardized (scores between 0.9 and 0.85) and low (scores below 0.85). job descriptors as well as actual up-to-date job requirements Categorization of similarity scores was based on a calibration (that also provide additional detail—for example, the ‘software skill’ category as mentioned above). In a second step, results are A Future of Jobs for All 23

28 Table A2: Detailed information on scaling and weighting of inputs for calculation of similarity scores Weighting for Type of information for scaling Input Scaling Definition similarity score Skills learned through Knowledge Level 0–7 education/training/experience 1 (equal weighting Learning acquired through practice of knowledge, Level KSA measure 0–7 and experience, practice used to Skills skills and facilitate knowledge acquisition abilities within O*NET KSA measure) Similar job activities and behaviors data Level Abilities 0–7 which underlie the work functions Level 0–7 1 Work Activities Tasks required to perform the role Requirements for each occupation Distribution 0 –100 1 Education, Training and Experience 1 by education and work experience Common, non-specialized skills required by job applicants Percent of job postings to be considered for the role 0 –100 Baseline skills containing skill name 1 (applicable to broad categories (equal weighting of jobs) of baseline, specialized Skills measure Skills particular to industry Percent of job postings Specialized or occupation, not easily and software 0 –100 containing skill name skills transferable skills within skills measure) BGT Skills related to the use, design, Percent of job postings data maintenance and repair of Software skills 0 –100 containing skill name software Percent of job postings Year of experience required for 0 –100 containing experience Experience the role requirement Education x 1 Experience Percent of job postings Years of education (and type: 0 –100 AA, BA, MA, PhD) required for containing educational Education the role requirement Source: Burning Glass Technologies. Categories of work experience are measured in time ranges and include: On-Site of In-Plant Training, On-the-Job Training, Related Work Experience and Required Level of Note: Education. Required Level of Education is measured in types of educational qualifications, including high school diploma, associate’s degree, bachelor’s degree and others. The final measure here indicates the occupation’s distribution across joint time and education/experience requirements for each occupation by either educational requirement-type (for example, Required Level of Education-Bachelor’s Degree) or work experience type-time requirement (for example, On-the-Job Training-over 6 months, up to and including 1 year). combined into a joint similarity score by calculating a weighted — O*NET occupational data: In a first step, the similarity score is calculated for ‘Knowledge’, ‘Skills’, and ‘Abilities’ (‘KSA’) as average between BGT and O*NET similarity scores. a group. In a second step, the similarity score is calculated Individual similarity scores for Burning Glass Technologies 3 for ‘Work Activities’ and ‘Education/Training/Experience’. job postings data and data from O*NET are computed by In a third step, a weighted average of similarity scores for calculating the similarity of requirement profiles for each separate pair of jobs. This is done by using a technique known as cosine KSA, Work Activities and Education/Training/Experience is 1 calculated (see similarity. Table A 2 for technical definitions of categories, The features of every job can be expressed in the form of a scalings and weightings). vector, which consists of the skill demand frequency, education, In a Burning Glass Technologies job postings data: — 2 Two jobs can then be compared and experience requirements. first step, the similarity score is calculated for different ‘Skill by calculating the similarity score between their respective Clusters’ (including ‘Baseline’, ‘Specialized’ and ‘Software’ vectors. An identical pair of jobs would have identical vectors of skills). In a second step, the similarity score is calculated features, and hence a similarity score of 1. The more different for ‘Experience’ and ‘Education’. In a third step, a weighted a pair of jobs, the closer their similarity score is to 0. Similarity average of similarity scores for measures for experience, scores between vectors of characteristics for jobs are calculated for technical Table A 2 education, and skills is calculated (see as follows: definitions of categories, scalings and weightings). Towards a Reskilling Revolution 24

29 Table A3: Example of an O*NET job zone: Job zone three (of five): medium preparation needed Education Most occupations in this zone require training in vocational schools, related on-the-job experience or an associate's degree. Related experience Previous work-related skill, knowledge or experience is required for these occupations. For example, an electrician must have completed three or four years of apprenticeship or several years of vocational training, and often must have passed a licensing exam, in order to perform the job. Job training Employees in these occupations usually need one or two years of training involving both on-the-job experience and informal training with experienced workers. A recognized apprenticeship program may be associated with these occupations. Job zone examples These occupations usually involve using communication and organizational skills to coordinate, supervise, manage or train others to accomplish goals. Examples include hydroelectric production managers, travel guides, electricians, agricultural technicians, barbers, nannies and medical assistants. Source: US Bureau of Labor Statistics. Utilizing data from the US Bureau of Labor Statistics is Condition 2: Job transition does not require huge leaps in beneficial for our analysis as the data set contains comprehensive education and experience and widely acknowledged information on employment on an When assessing job transitions, similarity scores are the main but individual job-level for the United States, aligned to our job not the only way of assessing viable job paths. Other elements categorization taxonomy. Connecting US Bureau of Labor that we take into account are the level of education (i.e. the formal Statistics employment data (based on SOC codes) to job mechanism for acquiring skills and knowledge) required and the postings data from Burning Glass Technologies and O*NET data level of experience (i.e. forming and improving skills to apply a (both based on O*NET codes) is achieved via the official O*NET- given knowledge) required, both as measured in years. SOC 2010 to SOC 2010 crosswalk. Where there was more than O*NET uses a classification known as ‘job zone’ which one O*NET code for a given SOC code, employment numbers incorporates these measures into each occupation. There are (2016 and 2026) were distributed to O*NET codes according five job zone categories. Any two occupations that are within the to proportions derived from the distribution of the number same job zone are similar in terms of the amount of education of job postings by O*NET code provided by Burning Glass required to do the work, how much related experience is required Technologies. to do the work, and how much on-the-job training is required to For the purpose of identifying viable job transition options, do the work. An example of a job zone from O*NET’s definition is we exclude job transitions that would involve transitions to jobs shown in Table A3 . that are expected to decline by 2026 in the US Bureau of Labor To avoid huge leaps in education and experience Statistics projections. requirements for two jobs, we exclude job transition options to job zones that are more than one job zone up or down from the starting job when identifying viable job transition options. Condition 4: Job transition opportunity leads to a level of wage continuity (or increase) that allows individuals to maintain (or improve) their present standard of living Condition 3: Job transition opportunity involves moving to target jobs that are not expected to decline in number When assessing opportunities for job transitions, one of the key desirable conditions is that living standards of the individual Assessing viable job transition options requires taking into do not decrease after the transition to the new job. This is best account the long-term sustainability of these job transition moves. assessed by the comparison of wage levels between the starting Figures on current employment and expected employment per and subsequent job, and the preference is for this to remain job reveal which jobs might present viable employment options stable or increase after the job transition. for workers in the future and which jobs are expected to decline in number. In this report, we use data on employment in 2016 and expected employment in 2026 from the US Bureau of Labor Statistics. A Future of Jobs for All 25

30 Table A4: Optimization conditions for Job Transition Model Constraints Utility function There are no job transitions to jobs with lower wages The sum of job transitions with each job transition, weighted by corresponding sum of similarity There are only job transitions from jobs where expected employment in 2026 is lower than in 2016 score and normalized percentage wage increase (between zero and There are no job transitions to jobs where expected employment in 2026 is lower than in 2016 one) There are no job transitions from jobs in job zone 5 (this is because job zone 5 comprises jobs such as CEOs, managers and scientists, where simulation of job transitions yield unlikely results) There are no job transitions with a similarity score of less than 0.85 Only job transitions to jobs in one job zone lower, equal or one job zone higher are feasible Employment per job is smaller than or equal to projected future employment in 2026 from the job postings data. The calculation logic of additional Job Transition Pathway Optimization Model experience and education required depends on whether a job Leadership lens transition happens within the same job family or between different The leadership lens perspective utilizes a job transition model, job families. based on the viability and desirability conditions set out above, Within a job family, additional experience/education required to simulate job movements using a Linear Programming Model is calculated by subtracting the average experience/education in that maximizes the value of a utility function and is restricted by a the starting job from the average experience/education required Table A4 provides an overview of the certain set of constraints. in the target job. Only positive differences are considered (i.e. utility function and constraints. cases where experience/education requirements in the target job As a basis, the job transition viability and desirability criteria are higher than in the starting job). The underlying assumption set out above are included in the constraints in the main model is that workers can fully transfer their experience/education to in this section. We have limited the optimization to constrain job different jobs in the same job family. transitions to jobs where there is no fall in wages (in relation to For job transitions between job families, we differentiate their starting point). We use Rglpk_solve_LP() in R to arrive at between experience and education. With experience, we assume a solution that takes as its main constraint the number of jobs that the additional experience required is the full amount of in the US economy in 2016 and 2026 and looks to place all average experience required in the target job. The underlying employees who are displaced into growing job families. assumption is that workers cannot transfer experience to jobs The number of job transitions by gender is calculated in different job families. With education, we assume that the by multiplying the total number of job transitions with the additional education required is the full amount of education proportion of women-to-men in each starting job. The underlying required in the target job. (However, we correct for high school assumption is that the distribution of gender across workers education—12 years are included in number of years of average transitioning to new jobs is equal to the distribution of gender in education required—as we assume that workers do not have to the starting jobs. repeat their high school education, even if they transition to jobs within a different job family.) Individual lens For all of these options, the average wage increase is calculated by subtracting the average wage in the target job from The individual lens perspective shows the job transition options the average wage in the starting job. The average wages are available to workers in any given occupation. It restricts those based on job postings data. options to those that meet the viable job transition criteria outlined above. In this section of the report, selected illustrative jobs are Notes shown, together with their viable job transition options. Further 1 https://en.wikipedia.org/wiki/Cosine_similarity. sub-sets of these job transition options are shown, restricting 2 https://en.wikipedia.org/wiki/Feature_vector. these potential job opportunities according to additional criteria. The rationale behind this step-wise calculation is to ensure that the effects 3 (A fuller set of such job transition pathways is also shown in of education, training and experience are not underweighted. If these Appendix B: Job Transition Pathways.) factors were to be merged directly with the knowledge, skills, and abilities component, their effects would be diluted, since the O*NET taxonomy The amount of additional experience and education required contains many more categories of skills than categories of education, training to facilitate a job transition is calculated using information on and experience. average experience and education required to perform a job Towards a Reskilling Revolution 26

31 Appendix B: Examples of Pathways Figure B1: Examples of Pathways for Secretaries and Administrative Assistants Insurance Claims Clerks Office and Administrative Occupations 44 wage: $41,000 opportunities similarity score: 0.86 with pay rise Library Assistants, Clerical Office and Administrative Occupations 8 wage: $27,000 Secretaries and opportunities similarity score: 0.89 Administrative with pay cut Assistants Office and Administrative Logisticians Production, Planning & Expediting Clerks Occupations Business and Financial Operations Occupations Office and Administrative Occupations wage: $36,000 wage: $49,000 wage: $78,000 similarity score: 0.91 similarity score: 0.92 Concierges Recycling Coordinators Transportation Occupations Personal Care and Service Occupations wage: $50,000 wage: $31,000 similarity score: 0.89 similarity score: 0.90 Job Job family Key Remuneration Similarity score with previous job Burning Glass Technologies and US Bureau of Labor Statistics. Source data: A Future of Jobs for All 27

32 Figure B2: Examples of Pathways for Cashiers Reservation and Transportation Ticket Agents and Travel Clerks 34 Office and Administrative Occupations opportunities wage: $38,000 with pay rise similarity score: 0.92 Hosts and Hostesses, Restaurant, Lounge and Coffee Shop 4 Food Preparation and Serving Occupations opportunities wage: $21,000 with pay cut Cashiers similarity score: 0.93 Sales and Related Occupations First-Line Supervisors of Retail Salespersons wage: $22,000 Sales and Related Occupations Retail Sales Workers wage: $27,000 Sales and Related Occupations similarity score: 0.94 wage: $44,000 similarity score: 0.92 Baristas Food Service Managers Food Preparation and Serving Occupations Food Preparation and Serving Occupations wage: $21,000 wage: $56,000 similarity score: 0.95 similarity score: 0.86 Figure B3: Examples of Pathways for Bookkeeping, Accounting & Auditing Clerks Brokerage Clerks Office and Administrative Occupations 14 wage: $52,000 opportunities similarity score: 0.88 with pay rise Library Assistants, Clerical Office and Administrative Occupations 6 wage: $27,000 Bookkeeping, opportunities similarity score: 0.85 Accounting and with pay cut Auditing Clerks Office and Administrative Eligibility Interviewers, Title Examiners, Abstractors Occupations Government Programs and Searchers wage: $40,000 Office and Administrative Occupations Business and Financial Operations Occupations wage: $51,000 wage: $44,000 similarity score: 0.95 similarity score: 0.88 Paralegals and Legal Assistants Court Clerks Business and Financial Operations Occupations Office and Administrative Occupations wage: $53,000 wage: $39,000 similarity score: 0.91 similarity score: 0.86 Job Job family Key Remuneration Similarity score with previous job Burning Glass Technologies and US Bureau of Labor Statistics. Source data: Towards a Reskilling Revolution 28

33 Figure B4: Examples of Pathways for Assembly Line Workers Rail Car Repairers Installation, Maintenance and Repair Occupations 59 wage: $54,000 opportunities similarity score: 0.89 with pay rise Packers and Packagers, Hand Transportation Occupations 23 wage: $24,000 opportunities similarity score: 0.89 Assembly with pay cut Line Workers Production Occupations First-Line Supervisors of Construction Construction Labourers wage: $33,000 Construction and Extraction Occupations Trades and Extraction Workers wage: $38,000 Construction and Extraction Occupations similarity score: 0.88 wage: $68,000 similarity score: 0.87 Animal Breeders Nursery Workers Farming, Fishing and Forestry Occupations Farming, Fishing and Forestry Occupations wage: $42,000 wage: $24,000 similarity score: 0.87 similarity score: 0.87 Figure B5: Examples of Pathways for Customer Service Representatives Real Estate Sales Agents Sales and Related Occupations 31 wage: $60,000 opportunities similarity score: 0.86 with pay rise Baggage Porters and Bellhops Personal Care and Service Occupations 29 wage: $25,000 opportunities Customer Service similarity score: 0.90 with pay cut Representatives Office and Administrative Occupations Credit Counselors Insurance Claims Clerks Office and Administrative Occupations Business and Financial Operations Occupations wage: $35,000 wage: $49,000 wage: $41,000 similarity score: 0.89 similarity score: 0.87 Solar Sales Representatives Retail Salespersons Sales and Related Occupations and Assessors wage: $27,000 Sales and Related Occupations similarity score: 0.90 wage: $93,000 similarity score: 0.89 Job Job family Key Remuneration Similarity score with previous job Burning Glass Technologies and US Bureau of Labor Statistics. Source data: A Future of Jobs for All 29

34 Figure B6: Examples of Pathways for Heavy and Tractor-Trailer Truck Drivers Crane and Tower Operators Transportation Occupations 31 wage: $55,000 opportunities similarity score: 0.85 with pay rise Baggage Porters and Bellhops Personal Care and Service Occupations 48 wage: $27,000 opportunities similarity score: 0.90 Heavy and with pay cut Tr actor-Tr ailer Truck Drivers Transportation Gas Plant Operators Sailors and Marine Oilers wage: $44,000 Production Occupations Transportation Occupations wage: $46,000 wage: $68,000 similarity score: 0.86 similarity score: 0.96 Rotary Drill Operators, Painting, Coating and Decorating Workers Oil and Gas Production Occupations Construction and Extraction Occupations wage: $57,000 wage: $32,000 similarity score: 0.87 similarity score: 0.88 Figure B7: Examples of Pathways for Travel Agents Police, Fire and Ambulance Dispatchers 23 Office and Administrative Occupations opportunities wage: $41,000 with pay rise similarity score: 0.85 Hotel, Motel and Resort Desk Clerks 31 Office and Administrative Occupations opportunities wage: $24,000 with pay cut Travel Agents similarity score: 0.90 Sales and Related Occupations Real Estate Sales Agents Real Estate Brokers wage: $40,000 Sales and Related Occupations Sales and Related Occupations wage: $79,000 wage: $59,000 similarity score: 0.94 similarity score: 0.87 Court Reporters Travel Guides Personal Care and Service Occupations Business and Financial Operations Occupations wage: $36,000 wage: $57,000 similarity score: 0.90 similarity score: 0.87 Job Job family Key Remuneration Similarity score with previous job Burning Glass Technologies and US Bureau of Labor Statistics. Source data: Towards a Reskilling Revolution 30

35 Figure B8: Examples of Pathways for Construction and Building Inspectors Gas Plant Operators Production Occupations 6 wage: $68,000 opportunities similarity score: 0.85 with pay rise Construction Labourers Construction and Extraction Occupations 6 wage: $38,000 Construction opportunities similarity score: 0.87 and Building with pay cut Inspectors Construction and Extraction Nuclear Monitoring Technicians First-Line Supervisors of Construction Occupations Life, Physical and Social Science Occupations Trades and Extraction Workers wage: $78,000 wage: $61,000 Construction and Extraction Occupations similarity score: 0.86 wage: $68,000 similarity score: 0.89 Air Traffic Controllers Traffic Technicians Transportation Occupations Transportation Occupations wage: $49,000 wage: $118,000 similarity score: 0.88 similarity score: 0.88 Figure B9: Examples of Pathways for Floral Designers Travel Guides Personal Care and Service Occupations 6 wage: $36,000 opportunities similarity score: 0.87 with pay rise Gaming Change Persons and Booth Cashiers 2 Sales and Related Occupations opportunities Floral Designers wage: $26,000 with pay cut similarity score: 0.86 Arts, Design, Entertainment, Sports and Media Occupations Talent Directors Makeup Artists, Theatrical Arts, Design, Entertainment, Sports and Media and Performance wage: $28,000 wage: $94,000 Personal Care and Service Occupations similarity score: 0.90 wage: $72,000 similarity score: 0.87 Public Address System Hosts and Hostesses, Restaurant, Lounge and Coffee Shop and Other Announcers Arts, Design, Entertainment, Sports and Media Food Preparation and Serving Occupations wage: $21,000 wage: $43,000 similarity score: 0.87 similarity score: 0.85 Job Job family Key Remuneration Similarity score with previous job Burning Glass Technologies and US Bureau of Labor Statistics. Source data: A Future of Jobs for All 31

36 Figure B10: Examples of Pathways for Radio and Television Announcers Program Directors Arts, Design, Entertainment, Sports and Media 29 Occupations opportunities wage: $94,000 with pay rise similarity score: 0.90 Umpires, Referees and Other Sports Officials 6 Radio and Arts, Design, Entertainment, Sports and Media opportunities Television wage: $36,000 with pay cut Announcers similarity score: 0.86 Arts, Design, Entertainment, Sports and Media Makeup Artists, Theatrical Music Composers and Arrangers Occupations Arts, Design, Entertainment, Sports and Media and Performance Occupations Personal Care and Service Occupations wage: $48,000 wage: $61,000 wage: $72,000 similarity score: 0.87 similarity score: 0.89 Subway and Streetcar Operators Radio Operators Transportation Occupations Arts, Design, Entertainment, Sports and Media Occupations wage: $62,000 wage: $47,000 similarity score: 0.85 similarity score: 0.85 Figure B11: Examples of Pathways for Buyers and Purchasing Agents, Farm Products Business Continuity Planners Business and Financial Operations Occupations 70 wage: $75,000 opportunities similarity score: 0.85 with pay rise Loan Counselors Business and Financial Operations Occupations 37 Buyers and wage: $50,000 opportunities Purchasing similarity score: 0.87 with pay cut Agents, Farm Products Business and Financial Operations Online Merchants Web Administrators Occupations Computer and Mathematical Occupations Business and Financial Operations Occupations wage: $89,000 wage: $75,000 wage: $64,000 similarity score: 0.85 similarity score: 0.88 Nursery and Greenhouse First-Line Supervisors of Agricultural Managers Crop and Horticultural Workers Farming, Fishing and Forestry Occupations Farming, Fishing and Forestry Occupations wage: $76,000 wage: $49,000 similarity score: 0.90 similarity score: 0.87 Job Job family Key Remuneration Similarity score with previous job Burning Glass Technologies and US Bureau of Labor Statistics. Source data: Towards a Reskilling Revolution 32

37 Figure B12: Examples of Pathways for Postmasters and Mail Superintendents 0 opportunities with pay rise Police, Fire and Ambulance Dispatchers 28 Office and Administrative Occupations Postmasters opportunities wage: $41,000 and Mail with pay cut similarity score: 0.86 Superintendents Community and Social Service Occupations wage: $72,000 Reservation and Transportation Customs Brokers Business and Financial Operations Occupations Ticket Agents and Travel Clerks wage: $75,000 Office and Administrative Occupations similarity score: 0.86 wage: $38,000 similarity score: 0.87 Figure B13: Examples of Pathways for Computer Programmers Computer Systems Analysts Computer and Mathematical Occupations 18 wage: $92,000 opportunities similarity score: 0.95 with pay rise Web Developers Computer and Mathematical Occupations 6 wage: $72,000 opportunities Computer similarity score: 0.92 with pay cut Programmers Computer and Mathematical Occupations Automotive Engineers Software Quality Assurance Architecture and Engineering Occupations Engineers and Testers wage: $85,000 wage: $90,000 Computer and Mathematical Occupations similarity score: 0.89 wage: $89,000 similarity score: 0.91 Telecommunications Network and Computer Systems Administrators Engineering Specialists Computer and Mathematical Occupations Computer and Mathematical Occupations wage: $104,000 wage: $85,000 similarity score: 0.88 similarity score: 0.89 Job Job family Key Remuneration Similarity score with previous job Burning Glass Technologies and US Bureau of Labor Statistics. Source data: A Future of Jobs for All 33

38 Figure B14: Examples of Pathways for Cooks, Fast Food Food Cooking Machine Operators and Tenders 73 Production Occupations opportunities wage: $30,000 with pay rise similarity score: 0.86 Combined Food Preparation and Serving Workers, incl. Fast Food 1 Food Preparation and Serving Occupations opportunity wage: $20,000 Cooks, Fast Food with pay cut similarity score: 0.94 Food Preparation and Serving Occupations Waiters and Waitresses First-Line Supervisors of Food wage: $21,000 Food Preparation and Serving Occupations Preparation and Serving Workers wage: $24,000 Food Preparation and Serving Occupations similarity score: 0.94 wage: $35,000 similarity score: 0.89 Combined Food Preparation and Refuse and Recyclable Material Collectors Serving Workers, including Fast Food Food Preparation and Serving Occupations Transportation Occupations wage: $20,000 wage: $38,000 similarity score: 0.88 similarity score: 0.94 Figure B15: Examples of Pathways for Mine Cutting and Channeling Machine Operators Structural Iron and Steel Workers Construction and Extraction Occupations 18 wage: $56,000 opportunities similarity score: 0.86 with pay rise Tile and Marble Setters Construction and Extraction Occupations 20 Mine Cutting wage: $45,000 opportunities and Channeling similarity score: 0.86 with pay cut Machine Operators Construction and Extraction Rail-Track Laying and Maintenance Subway and Streetcar Operators Occupations Transportation Occupations Equipment Operators wage: $62,000 Construction and Extraction Occupations wage: $51,000 similarity score: 0.86 wage: $53,000 similarity score: 0.86 Excavating and Loading Machine Nuclear Equipment and Dragline Operators Operation Technicians Transportation Occupations Life, Physical and Social Science Occupations wage: $78,000 wage: $45,000 similarity score: 0.91 similarity score: 0.85 Job Job family Key Remuneration Similarity score with previous job Burning Glass Technologies and US Bureau of Labor Statistics. Source data: Towards a Reskilling Revolution 34

39 System Initiative Partners The World Economic Forum would like to thank the Partners of the System Initiative on Shaping the Future of Education, Gender and Work for their support and guidance of the System Initiative. Infosys • A.T. Kearney • AARP JLL • • Johnson Controls • • Accenture Lego Foundation Adecco Group • • • Limak Holding • African Rainbow Minerals LinkedIn Alghanim Industries • • ManpowerGroup • • AlixPartners • Bahrain Economic Development Board Mercer (MMC) • Bank of America • • Microsoft Corporation • Nestlé Barclays • Nokia Corporation Bill and Melinda Gates Foundation • • • • NYSE Bloomberg Omnicom Group • • Booking.com Ooredoo • • Boston Consulting Group PayPal Centene Corporation • • Centrica Pearson • • Chobani Procter and Gamble • • • • Dentsu Aegis Network PwC Dogan Broadcasting • Salesforce • SAP EY • • GEMS Education • • Saudi Aramco • Skanska AB • Genpact International • Google Tata Consultancy Services • GSK • • The Rockefeller Foundation • TupperwareBrands Corporation • Guardian Life Insurance HCL Technologies • • Turkcell Heidrick & Struggles • UBS • • Hewlett Packard Enterprise • Unilever • Home Instead • Willis Towers Watson • HP Inc. • Workday • Hubert Burda Media WPP • In addition to our Partners, the leadership of the System Initiative on Shaping the Future of Education, Gender and Work includes leading representatives of the following organizations: Council of Women World Leaders; Endeavor; Haas School of Business, University of California, Berkeley; International Labour Organization (ILO); International Trade Union Confederation (ITUC); JA Worldwide; London Business School; MIT Initiative on the Digital Economy; Ministry for Education of the Government of Singapore; Ministry for Employment of the Government of Denmark; Ministry of Employment, Workforce Development and Labour of the Government of Canada; Department for Planning, Monitoring and Evaluation of the Presidency of South Africa; Office of the Deputy Prime Minister of the Russian Federation; The Wharton School, University of Pennsylvania; and United Way Worldwide. To learn more about the System Initiative, please refer to the System Initiative website: https://www.weforum.org/system-initiatives/shaping-the-future-of-education- gender-and-work. A Future of Jobs for All 35

40

41 Acknowledgements Towards a Reskilling Revolution: A Future of Jobs for All is an insight tool published by the World Economic Forum’s System Initiative on Shaping the Future of Education, Gender and Work, in collaboration with The Boston Consulting Group and using proprietary data provided exclusively for this report by Burning Glass Technologies. PROJECT TEAM At the World Economic Forum At The Boston Consulting Group Saadia Zahidi Rainer Strack Senior Partner and Managing Director Head of Education, Gender and Work; Member of the Executive Committee Theodore Roos Principal Vesselina Ratcheva Seconded to the World Economic Forum Data Lead, Education, Gender and Work Nikolaus von Blazekovic Till Alexander Leopold Project Lead, Education, Gender and Work Consultant DATA PARTNER Burning Glass Technologies Dan Restuccia Chief Product and Analytics Officer Bledi Taska Chief Economist Soumya Braganza Senior Research Analyst Rachel Neumann Senior Research Analyst Thank you to those at The Boston Consulting Group who provided support and expertise: Hans-Paul Bürkner, Natalia Schmidt, Judith Wallenstein and Zheng Wei Yap. Thank you to the System Initiative on Education, Gender and Work team: Nada Abdoun, Piyamit Bing Chomprasob, Genesis Elhussein, Sofia Michalopoulou, Paulina Padilla Ugarte, Valerie Peyre, Brittany Robles, Pearl Samandari, Lyuba Spagnoletto and Susan Wilkinson. Thank you to Michael Fisher for his excellent copyediting work, Kamal Kamaoui and the World Economic Forum’s Publications team for their invaluable contribution to the production of this White Paper; and Neil Weinberg for his superb graphic design and layout. A Future of Jobs for All 37

42 The World Economic Forum, committed to improving the state of the world, is the International Organization for Public-Private Cooperation. The Forum engages the foremost political, business and other leaders of society to shape global, regional and industry agendas. World Economic Forum 91-93 route de la Capite CH-1223 Cologny/Geneva Switzerland +41 (0) 22 869 1212 Tel +41 (0) 22 786 2744 Fax [email protected] www.weforum.org

Related documents