FRIEDRICH-EBERT-STIFTUNG – Artificial Intelligence and Auto­mation in Retail MODERATE POLICY RESPONSE HIGH SEVERIT Y LOW Figure 2 Anticipated challenges and policy responses ANTICIPATED CHALLENGE Job transformation or redesign due to AI use Changing skills needs due to occupational restructuring Reduced work hours and/or income Displaced workers due to automation More frequent periods of unemployment Job polarisation and rising inequality Lack of social dialogue and Loss of employee voice and social dialogue due to workforce fragmentation Increased aggregate unemployment POLICY RESPONSE Responsive education system and better IAG Retraining and Lifelong learning (e.g. learner accounts) Minimum wage protection Social protection linked to individuals not jobs Active Labour Market Policies Profit sharing schemes Multi-level collective bargaining Robot tax Universal Basic Income or Citizens Dividend RADICAL of retail highlighted in this report, continuous professional development(CPD) and reskilling programs would remain desirable as they are intimately tied to job quality and career progression. This raises the question of who(or which mix of actors) should bear the burden of cost for training provision and certification – workers, firms, governments, or unions? Since the sector is dominated by price competition, few employers are likely to voluntarily engage in extensive staff training and skill development programs in spite of the benefits this can bring. 39 National contexts bear heavily on the kinds of out­comes possible. Participants pointed to several examples of agreements involving the state, unions and employers in the Nordic countries. At perhaps the most progressive end of the spectrum, Sweden’s government(along with employers and unions organised into employment councils) is developing a EUR 1bn per year program for CPD and retraining including extensive ›sabbatical‹ periods for retraining paid at 80% of former salary. 40 In Denmark, the HK service-sector union engages in widespread member training initiatives aimed at meeting changing industry requirements, but also presses for the right to and funding for retraining and CPD to be em­bedded in collective agreements with employers. By contrast, participants in southern Europe and Turkey noted that both employers and governments there appear less forthcoming with funds and time allowances for reskilling and training, leaving unions alone to bear the cost. While there can be benefits in terms of recruitment and growth from unions providing training, it should be noted that such union-funded skilling initiatives thrive in economies with high union density and where unions assume responsibilities for administering social benefits. In contexts with low density and weak em­beddedness, take-up of training is likely to remain low and ineffectual as a means for pressuring employers to retain and redeploy labour toward higher-productivity areas. 41 Retail workers already possess many tacit, unaccredited skills. For this reason, it is unlikely that the sector will expe­rience mass technological unemployment in the short- to medium-term. Building occupational identity and fighting for(existing) skills recognition and accreditation was high­lighted as priority by focus group participants. Unions should pressure both employers and regulators to establish sectoral training initiatives which provide workers with recognised qualifications or accreditation for existing skills. 14