CAMBRIDGE – Conventional economics has always had a blind spot when it comes to jobs. The problem goes back to Adam Smith, who placed consumers, rather than workers, on the throne of economic life. What matters for well-being, he argued, is not how or what we produce, but whether we can consume our preferred bundle of goods and services.

Modern economics has since codified this approach by capturing individual well-being in the form of a preference function defined over our consumption bundle. We maximize “utility” by selecting the goods and services that offer us the most satisfaction. Though each consumer is also a worker of some kind, jobs enter the equation only implicitly through the income they provide, by determining how much money we have available to spend on consumption.

Yet the nature of one’s job has implications far beyond one’s budget. Jobs are a source of personal dignity and social recognition. They help define who we are, how we contribute to society, and the esteem that society in turn bestows on us. We know that jobs matter because people who lose them tend to experience large and persistent reductions in life satisfaction. The monetary equivalent of such drops is typically a multiple of a person’s income, rendering compensation through government transfers (such as unemployment insurance) infeasible for all practical purposes.

More broadly, jobs are the cement of social life. When decent, middle-class jobs disappear – owing to automation, trade, or austerity policies – there are not just direct economic effects, but also far-reaching social and political ones. Crime rises, families break apart, addiction and suicide rates soar, and support for authoritarianism increases.

When economists and policymakers think of social justice, they typically focus on the “distributive” variety – who gets what? But, as political philosopher Michael J. Sandel argues, perhaps a more important yardstick is “contributive justice,” which refers to the opportunities to win the social respect that comes with good jobs and “producing what others need and value.”

While these issues are typically considered in the context of advanced economies, they are equally important for developing countries. In a rich country, a good job might be defined as one that enables a path to middle-class living standards, and that upholds core labor rights such as safe working conditions, collective bargaining, and regulations against arbitrary dismissal. In a poor country, a good job is typically one that simply affords a higher living standard than unproductive, back-breaking subsistence agriculture or a precarious existence in the informal sector.

In fact, people moving from bad jobs to better jobs encapsulates the entire process of structural change that drives economic development. Unlocking this process in a rapid, sustainable manner is crucial, and industrialization, historically, has been the main engine for doing so.

The trouble now is that manufacturing industries are no longer the labor-absorbing sectors they once were. A combination of factors – particularly the increased skill- and capital-intensity of modern manufacturing methods and stiff international competition to join global value chains – has made it very difficult for developing economies to increase employment in formal manufacturing. Even countries with strong industrial sectors – not least China – are experiencing declines in manufacturing as a share of total employment.

The inevitable consequence of these trends is that the bulk of the better jobs will have to be generated in services, both in developing and developed countries. But since most services in developing countries are highly unproductive and informal, this shift poses a major challenge. Making matters worse, most governments are unaccustomed to thinking of service sectors as growth engines. Growth policies – whether they relate to research and development, governance, regulation, or industrial policies – typically target large manufacturing firms that compete on world markets.

Difficult though it may be, governments must learn how to enhance productivity and employment simultaneously in labor-intensive service sectors. That means adopting measures with many of the same features of “modern industrial policy,” whereby the state, in exchange for job creation, pursues close, iterative collaboration with firms to remove obstacles to their expansion.

There are already some examples of this model around the world. Consider the Indian state of Haryana’s partnership (begun in 2018) with the ride-hailing services Ola and Uber. Established with the goal of increasing employment for young people by making it easier for these firms to identify and hire drivers, this public-private partnership is based on a clear quid pro quo. Haryana has eased regulations that hampered the services’ growth, shared databases of unemployed youth, and held exclusive job fairs for the companies, which in turn have made (soft) commitments to employ a meaningful number of young people.

The agreement is dynamic. Allowing for the terms to adapt to changing circumstances helps to build mutual trust without binding the firms to rigid conditionalities. In less than a year, the partnership has created more than 44,000 new jobs for Haryana’s youth.

Of course, services are a hodge-podge of different activities, with great heterogeneity in the size and shape of firms. Any realistic program to expand productive employment in services will have to be selective, focusing on those firms and subsectors that are more likely to be successful. There will need to be experimentation, and local governments – municipalities and sub-national authorities – will often be in a better position than national officials to carry out pilot programs.

Ultimately, economic growth and equity both require a jobs-centered approach to development. While economic growth is possible only if workers move toward better, more productive jobs, equity requires improvements in the employment prospects for workers at the bottom of the income distribution. A growing middle class will then, in turn, boost domestic demand and reinforce job creation in services.

A services-based model cannot generate the kind of growth miracles that export-oriented industrialization produced in the past. But it can still lead to higher-quality growth, with much greater social inclusion and a broader middle class.

Dani Rodrik, Professor of International Political Economy at Harvard Kennedy School, is President of the International Economic Association and the author of Straight Talk on Trade: Ideas for a Sane World Economy (Princeton University Press, 2017).