Applying the Model to Good Jobs

The concept of “good jobs,” like clean water, is imprecise and needs to be op­erationalized in a way that is both evolving and context-dependent. Reason­able, attainable targets for the creation of good jobs must remain provisional, to be revised under new information.

Simplifying greatly, the uncertainty government agencies face in the ARPA-E case is principally about technological feasibility. The uncertainty in the dairy case derives largely from local adaptation. A “good jobs” program faces uncertainties of both kinds. Creating or preserving good jobs in a par­ticular place often depends partly on extending technological capabilities: mastering techniques that are wholly novel (at least in some particular ap­plication) or so new to a given locale that they must almost be reinvented to be mastered. Here ARPA-E's experience with active project management and collaborative review and adjustment of milestones is directly relevant. But fostering good jobs depends at least as much on solving highly idiosyncratic, place-specific problems: failures of coordination between local firms and training institutions and between firms and their (potential) supply-chain partners, and the managerial breakdowns or skill gaps within individual firms and institutions to which the coordination problems point.

Building a GoodJobs Economy 81 the time comes, and provisionally—subject to correction—in accord with the precepts of the governance model itself.

An immediate question has to do with the timing, scale, and scope of the obligations (and penalty defaults) to be imposed on private firms. If there is a genuine good jobs externality, and a national or subnational mandate to address it, there is no reason, in principle, that the obligation to do so should not be applied immediately to all firms in the relevant jurisdiction. But as just noted, those obligations are inherently broad, open-ended, and at least initially, ill-defined. They would begin with the requirement to make plans to progress toward forms of organization and deployment of technology that in combination produce better jobs, and to make such plans in coordination with relevant peers and institutional partners. But this may well be a too draconian first step. Unless we assume extraordinary consensus in favor of addressing the jobs externality or a dangerously coercive state authority, we cannot really imagine the regulator imposing on all firms, or even all firms in certain sectors, the obligation to make such plans; and if we cannot imagine that, still less can we envisage penalty defaults for persistent failure to make good faith efforts to comply.

It is easier to imagine imposing such requirements and penalties on actors who volunteer to participate in government programs designed to achieve the same outcome and conferring benefits in the form of improved regula­tion, better coordination, extensive customized support services, or the like in return for participation.

A key benefit of these voluntary arrangements over the medium term is to develop an inventory of “good practices”—a repertoire of contextualization measures variously suited to a wide range of settings—that can eventually guide application of the good jobs strategy to a larger set of firms, cutting the costs and increasingly the chances for early successes of broader cover­age. Put differently, the initial, selective projects would serve as a pilot pro­gram for the new system of regulation, with the qualification that pilots are usually understood as practical tests of promising concepts, whereas in this

case their purpose would be more to identify and begin to refine promis­ing approaches under real-world conditions than subject them to definitive tests. As formal obligations are extended, the arrangements would come to resemble the European regulatory model, with a uniform requirement of par­ticipation but responses highly differentiated by locale. The need for con­textualized support for the less capable actors drawn into the system would grow apace.

An intermediate arrangement might also be possible. Firms might be asked to make a choice between participation in customized compacts for good jobs with public agencies and submitting to a fixed regulatory regime that imposes a common, universal set of benefits and obligations linked to job creation: for example, a schedule of tax incentives/penalties in return for an increase of x percent per annum in the number of employees at wages at or abovey percent of the local median, where the rate ofjob creation may be tied to the business cycle.

With these design principles and staging practicalities in mind, an indus­trial policy on good jobs could be introduced in four steps. First the gov­ernment commits in legislation or by other means to address the problem of bad jobs and no jobs as a constitutional externality that threatens the foundations of our democracy and requires for its solution concerted coop­eration between regulators, service providers, and private actors. The fram­ing legislation mandates regulators with relevant authority to put in place information-generating regimes that allow for standard setting and revision. The same legislation creates an interagency body to periodically review and prompt improvement of regulatory responses, and to resolve coordination problems arising from them, while also providing funds and authority for voluntary programs in anticipation of an eventual, step-wise extension of regulatory reach.

Regulators who currently have delegated authority for areas directly af­fecting job abundance and quality—vocation training, agricultural and man­ufacturing extension, standard setting, and the like—introduce, in a second step, innovation-inducing and contextualizing governance mechanisms where they are not already in place, anticipating the need for support services to help vulnerable actors comply with increasingly demanding requirements. The requirements can take different forms, including specific employment quantity targets and/or standards.

Where current regulatory authority does not reach, the government cre­ates volunteer, public-private programs to advance the frontiers of technol­ogy and organization, or—and of equal and perhaps greater importance— provide support services and perhaps subsidies to help firms bridge the gap between their current low-productivity/low-skill position and participation in the advanced sector.

Finally, conditional on the success of voluntary arrangements, the scope of these practices would gradually be made obligatory for nonparticipating firms, starting with requirements for submitting credible plans for improv­ing the quality and quantity of jobs they offer, along with their competitive position, by better organization and use of skill and technology. Where ap­propriate, plans should anticipate coordination with other firms and institu­tions. Penalty defaults would be imposed on laggard firms that, despite the availability of support services, persistently fail to comply.

To place our proposed framework in sharper relief, we discuss briefly how it relates to some existing initiatives for promoting manufacturing and job creation.

COMPARISON WITH CURRENT INITIATIVES

Of the three major components of the building good jobs program—exten­sion services and cooperative research programs for existing firms; job cre­ation and attraction policies; and active labor market policies or workforce development—it is to the last, workforce development, that the governance practices in advanced technology and European regulation (described previ­ously) have been most consistently applied with demonstrable success. In the case of job attraction, the experience has been nearly the reverse. Local and state politicians outbid each other to win outside investments in new facilities, more often than not in deals that (in contrast to contracts for in­novation) specify all terms of the exchange fully in advance: so much in sub­sidies for each job created at an agreed wage rate in a facility of an agreed type.

Many of the most successful workforce development programs trace back to Project QUEST (Quality Employment through Skills Training), founded in San Antonio in 1992,¹² in response to a wave of plant closings—an early portent of broader dislocations to come. The displaced workers lacked the skills for the new jobs being created in health care, IT, and other sectors; the service-sector jobs for which they were qualified paid too little to support a middle-class family. Two faith-based social movement organizations, seeing the urgent need for a program to equip the region's largely Hispanic popula­tion for good jobs, secured municipal funding to create Project QUEST (War­ren 2011).

The new project faced a double challenge. On the one hand, it had to iden­tify emerging opportunities on the local labor market, alert the city's com­munity college system (then still inattentive to business needs) to them, and help shape the substance and timing of new courses to meet the needs of firms and students. On the other hand, it had to learn to support a popula­tion of high-risk learners, almost all of whom needed to pass difficult reme­dial courses to qualify for further study, and many of whom had family and financial burdens on top of anxieties about returning to school.

In facing these challenges, Project QUEST turned to former military members with long experience in workforce development. The first executive director was the former commander of the Air Force Recruiting Service; his successor, and many managers later hired, had a similar background. These managers brought with them not habits of military discipline and hierarchy but rather the culture of continuous improvement—the continuous moni­toring of individual cases and rapid learning from disruptions at the core of our governance principles—that took root in many parts of the US military before it become standard operating procedure in much of the economy.¹³ An expression of this culture was the early creation of a dedicated management information system, highly unusual for an organization like Project QUEST at the time, to track the performance of individual students, both to keep their counselors abreast of their progress and to allow continuing review of overall organizational performance.

To be eligible to participate in QUEST, students must demonstrate need (generally earnings of less than 50 percent of the local median wage) and levels of literacy and numeracy sufficient to ensure reasonable chances of succeeding at the intense remedial programs typically needed to prepare for the required, basic courses in community college programs. Once admitted, students design in collaboration with a counselor a bundle of “wraparound” services and supports to help them surmount stumbling blocks on the path to completing training, including subsides for tuition, child care, or rent or services to address problems of transportation, health, or domestic violence.

Counseling is continuous and intense. Students meet their counselors in­dividually and in small, stable groups in weekly, hour-long sessions, where they share problems and devise mutual support strategies. A key purpose of these meetings is to identify and respond to emergent problems before they trigger a cascade of failure ending in withdrawal from the program. In effect, the counseling sessions in combination with information about students' class performance allows for continuous adjustment of the wraparound sup­port bundle.

A recent randomized controlled trial (RCT) evaluation of the earnings of QUEST participants nine years after leaving the program demonstrates the effectiveness of the approach (Roder and Elliott 2019). QUEST participants earn roughly 10 percent more per year than the control group, and the gap does not diminish—and may be growing—over time. Crucially, the differ­ence in earnings is the greatest for the most at-risk subgroups: students who took part in QUEST when they were older than the normal school-going age, with children and additional burdens.

In recent years, as community colleges are increasingly drawn into train­ing partnerships with local firms, more students from more diverse back­grounds seek new qualifications, and the failures of limited, “light touch” interventions to increase completion rates become more conspicuous, the schools themselves are successfully providing many of the individualized services originally offered by QUEST. A leading example is the Accelerated Study in Associate Programs (ASAP) of the City University of New York (CUNY).

Like Project QUEST, ASAP provides financial support (to bridge the gap between the available aid and tuition and other fees) and wraparound ser­vices, above all a dedicated adviser for each student who furnishes frequent and comprehensive support. Again, a principal goal is to identify and resolve issues before a student drops out of school (Weiss et al. 2019). Cumulatively, over the course of the three-year program, these customized interventions have produced a striking increase in completion rates. Nearly 40 percent of the students in the program group in an RCT study graduated by the end of the program; the graduation rate for control group students was 22 percent. What makes ASAP uniquely successful, the authors of the study find, “is that its multiple, integrated, and well-implemented services address multiple prevalent barriers to student success, and those services are offered for three full years” (Weiss et al. 2019, 279): in our terms, continuing contextualization based on continuous monitoring.

We would expect a program designed to adjust to local circumstance to be scalable, and ASAP is proving to be. Since 2014, three Ohio community colleges have implemented ASAP, and early impact assessments show results comparable to those obtained in New York (Sommo, Cullinan, and Manno 2018). Community college leaders are following the success of ASAP closely and devising their own systems of comprehensive support, with the goal, in the words of one, of “making help unavoidable.” In the most ambitious cases, growing confidence in the ability to train low-skill workers is encouraging community colleges to enter extensive partnerships with large firms such as Amazon, where the school offers customized training and student sup­port and the company pays tuition expenses, synchronizes the work schedule of participating employees to mesh with school needs, and, perhaps most crucially, provides a career ladder from unskilled work into management for those who complete the program.¹⁴

If workforce development programs are increasingly aware of the need for continuous learning in response to the uncertainties of context, programs that directly target employment creation by attracting inward investment seldom are. The most visible of such programs, typically administered by states rather than the federal government, are tax incentives provided to large investors in return for specific commitments on job creation. The Fox­conn and Amazon deals, in Wisconsin and New York, respectively, are recent high-profile examples. The Taiwanese company Foxconn had agreed to cre­ate 13,000 well-paying jobs in Wisconsin in return for more than $4.5 bil­lion in government incentives. Amazon promised creating 25,000 jobs over a decade in return from an incentive package from New York valued at nearly $3 billion. Both arrangements have blown up amidst controversy; their fail­ures are instructive in ways that demonstrate the superiority of the alterna­tive approach we are suggesting here.

Essentially, the Foxconn and Amazon deals—as well as similar tax incen­tive programs—were predicated on ex ante Contractibility (and hence a sta­ble environment). With enough predictability about market and technology conditions, firms can make rational calculations about employment commit­ments. And the states have the assurance that firms will deliver. Once the contract is written down, the state remains at arms' length from the firm. In Amazon's case, the company said it wanted cities to “think big.” In reality, as one commentator has noted, “the creative thinking was exclusively focused on incentive offers” (Jensen 2019). If the firm turns out to be unwilling or unable to carry out the terms of the contract—as was the case with Foxconn and Amazon, the former because of unforeseen changes in demand and tech­nology and the latter because of unexpected political fallout—there is little room for revision or renegotiation.

Bartik (2018, 2019) has studied such tax incentive programs more broadly and concludes that, even when they work, they are not very cost-effective. This is especially true when local incentives have to be financed by cuts in public expenditures elsewhere (e.g., education or infrastructure). Bartik ar­gues that the most effective employment programs focus specifically on lo­cal labor demand and supply conditions. He emphasizes three strategies in particular. The first—and, Bartik finds, by far the most cost-effective—is the provision of customized public services to small and medium-sized enter­prises. These include job training tailored to local employers and run by local community colleges, and “manufacturing extension services” that provide marketing and technology advice. The second is targeted investments in workers' skills and training, ranging from preschool programs to wage subsi­dies, and the third is infrastructure programs that increase land supply and thereby lower business costs.

All three strategies are consistent with our emphasis on iterative fine­tuning and evolving standard setting in lieu of ex ante rules. The design of locally effective incentive packages along these lines obviously requires extensive information discovery and trial and error on the part of local de­velopment agencies, heightening the importance of organizational arrange­ments of the type we have discussed here. Note also that while Bartik's (2018) focus is on manufacturing employment, our proposals would apply to service sectors as well. This is important since it is unlikely that the long-term, secu­lar decline in the share of manufacturing employment can be reversed.

More broadly, good practice in industrial policy has moved away from presumptive approaches that assume the government has a good fix on the underlying problem and the requisite solutions. For example, industrial parks of enterprise zones presume that the absence of good jobs is due to, say, high taxes and poor infrastructure, and they create spaces where neither is a problem. Such prepackaged solutions work poorly when firms face dif­ferentiated obstacles—lack of workers with appropriate skills or inadequate access to specialized technologies, for example. The collaborative framework we have outlined here has the advantage that it is explicitly diagnostic—that is, focused on information discovery.