Blog

In his stellar recent book, The Human Network, Matthew O. Jackson discusses reasons why the internet might cause erosion of news quality. First, the costs of starting up an outlet to disseminate news are lower than ever. In many circumstances, lower entry barriers are good for consumers—they spur competition in an industry, resulting in higher quality and lower costs. However, a balancing force in markets for information like news is that it is costly to determine the quality of the information received. When there are many providers, the market for news becomes more congested—assessing the quality of different sources of information is hard and takes time, and that effort and time gets wasted on low-quality or pernicious information providers. Past research suggests people are not good at distinguishing fact from fiction on the internet.¹

A second problem is that social media and news aggregators reduce the payoff from investing in investigative journalism. High quality reporting published on the internet may be immediately re-reported by other outlets and spread through social media feeds. To get basic, up-to-date  information about current events, people do not need to read the original source. In the past, this was not the case. 100 years ago, when a newspaper had a new scoop, it would take at least a day for other papers to disseminate that same information. Newspapers that produced an investigative piece had a big first mover advantage, and the desire to be first encouraged healthy competition between newspapers. Consumers wanted to subscribe to the newspapers that gave them quality news before it became stale. Today there is little first mover advantage for producing news. The speed of replication makes investigative journalism less lucrative, meaning less investigative journalism gets produced. The legwork is costly, and in the past few decades the benefit has often not been worth the cost, especially for local news. ²

These two phenomena can combine to erode the quality of news. Creating high quality news is hard, and pieces get replicated almost immediately; creating bad or fake news is easy; people are not very good at distinguishing between the different types, making the gap in revenue for both types of news inefficiently small. A countervailing effect is reputation. If people generally trust established news outlets more than less reputable ones, they may tend to still patronize such news sites more. Undermining the credibility of established outlets can diminish reputational advantages.

Good interventions to improve these possible issues are not obvious. Somehow increasing barriers to spreading information can potentially be very costly if it limits access to legitimate knowledge. While diversity of information might lead to congestion in news, it can also help people learn things, spurring innovation and making people better off in and of itself. Technology platforms have been grappling with the pitfalls of restricting information exposure in recent years. Algorithms developed by social media outlets to surface items on newsfeeds and filter out “bad” information get criticized for censoring views or for exerting excessive, unaccountable, black box control over what people see.

Similarly, encouraging the socially optimal level of investigative journalism is challenging and seems to contrast with frameworks for incentivizing R&D investment for many other types of goods. To incentivize costly investment in research and development, countries around the world have, for example, patent laws. These laws prevent people from immediately copying new inventions that were expensive to develop. Instituting patent or copyright laws for investigative journalism seems less tenable for many types of journalism. First, it may be impossible to enforce. Lawsuits against everyone tweeting about the contents of an article would be socially wasteful and restrict civil liberties. Further, there is compelling evidence that patent laws fail to boost innovation in the first place. Providing subsidies or prizes for investigative journalism is also tricky. Not all investigative journalism should be subsidized—especially fake or misleading news. Ideally pieces that are higher quality should be subsidized more, but measuring quality is hard. Readership is likely not a good proxy for quality, and further is hard to measure when people can get the same information from many different sources. Deciding what does and does not deserve to be subsidized could introduce unwanted bureaucracy, concentrated power, and bias in news. Coming up with mechanisms to encourage high-quality investigative journalism seems like an important area for research.

There are market failures in journalism today that likely have been magnified in the past several years. The easy creation and rapid spread of information around the web can reduce the stock of quality news. Less investigative journalism can reduce accountability and prevent diffusion of high-quality knowledge that spurs innovation. Thinking about how to incentivize quality journalism while preserving the rapid proliferation of information on the internet is an interesting research question.

This post is inspired by Fatima Farheen Mirza’s beautiful novel A Place for Us, which unpacks the lives of an Indian-American Muslim family in California. Each member of the family struggles in their own way with navigating their identity at the intersection of cultures and loyalties, at times clashing violently and at other times finding solace in each other. The central conflict is the inability of Amar, the youngest child and only boy, to feel a sense of belonging where he is born. His fall is a reminder that a comfortable material life cannot guarantee well-being and a fulfilled existence. A person’s community and how well they match with it matters for how happy they are, and it is not certain that economic growth and prosperity facilitate better matches.

Imagine that having loved ones matters a lot to a person’s well-being. Suppose there are two inputs to whether we love someone: our affinity for them, and how much time we spend with them. We develop loving relationships only by being around people who we are drawn to and spending enough time with them to connect emotionally. Neither of the inputs to a successful relationship can be purchased. There is only one resource we can use to foster loving relationships–our time. We can use our time to search for people who we might have an affinity with or to spend with potential loved ones in hopes of connecting more deeply.

Each person is endowed with a stock of potential loved ones when they are born: their family and the surrounding community. Suppose that similarity is a good predictor of affinity–we are more likely to feel a close connection to people who have similar personalities and life backgrounds. Then if the affinity between each pair of individuals is unknown beforehand, the endowment we are given at birth is arguably the most fortuitous possible one. We are matched with people who share our genes and our cultural context, which before our birth are the best predictors of similarity and affinity.

However, the quality of this endowment depends on the level of historical change. The lives of the people in our community do not have to be similar to our own. Suppose between the last generation’s birth and our own a major shock took place to the community. Shocks can be positive or negative. Perhaps our community experienced massive sudden economic growth, so our childhood is more comfortable than it was for the adults we grow up around. Alternatively, our community may have been involved in a devastating war before we were born, and we could be worse off than the past generation. Or perhaps we are the first generation to grow up in the internet age, and our community never previously experienced constant access to schoolmates or strangers on the web.

These shocks over time to the cultural context may reduce the affinity we have to the community we grow up in. To see this, instead imagine growing up centuries ago in a household of farmers. For many generations, our family has lived on the same plot of land and tilled the same fields. Our community has deeply ingrained cultural practices passed down for many years that everyone believes and adheres to. We all speak the same language and follow the same religion. No one thinks of leaving their birthplace to pursue opportunities elsewhere. In such a context, the quality of our endowment of relationships at birth must be quite high. The odds of finding people more similar in personality, ambition, and background are low.

In contrast, the intergenerational disruption in social conditions has been higher in recent centuries. Even the tools used by our community a generation ago (landlines, pagers, checkbooks, physical maps) are almost comically obsolete now. The way we speak, the way we play, and the education and occupation we choose for ourselves are all different. While our genes and upbringing still tie us to home, the people we feel most comfortable with may not be the ones who live near us.

A similar situation might hold for immigrant families and the way they experience and assimilate into a new cultural context. Zhou and Xiong (2005) discuss that second generation immigrants can try to adapt to their social environment in several different ways, including severing ethnic ties and adopting the dominant culture, adapting their native subculture in direct opposition to the dominant culture, or something in between. In the novel, none of these approaches quite works for Amar, leaving him isolated. Anecdotal evidence suggests such feelings may be common among second generation immigrants caught between cultures.

A lower quality endowment of relationships hurts us, and we have to face the tax in two ways in expectation. First, we have to invest time searching for people who have a high affinity with us and cultivating our relationships with them. Second, a lower quality endowment exposes us to more risk of not matching well with anyone. Perhaps we get unlucky, and all the acquaintances we come upon are not good fits for becoming loved ones. Maybe we were not quite born in the right time and place, and during our lives we fail to find people who we really love and care for and who return the feeling.

It is unclear whether economic growth helps remedy or exacerbates these problems. Growth can cause social disruption and be the source of a worsening endowment of acquaintances. Technological adoption, for instance, can erode intergenerational similarity. Further, unlike for material goods, economic growth does little to slacken our budget constraint for loving relationships–our time on earth stays relatively fixed. Life expectancies have been flat in developed countries in the past few decades, preventing us from looking indefinitely for people we match well with.

On the other hand, it may make it easier to find and keep in touch with people who have high affinity with us. For example, there is evidence online dating reduced search costs for romantic partners considerably. Technologies like messaging apps and Skype made it easier to talk to acquaintances who were far away, expanding the radius of people we might come to love. Still, it is uncertain how a lower quality endowment trades off against lower search costs in finding loved ones. Perhaps in the long run recommender systems will be so good that they can predict affinity even better than our natural endowment and match us with loved ones; it is not clear if or when that might happen, or whether it already has.

This is all to say that there is more to our well-being than economic growth, and the effect of systematic social change–even when driven by growing prosperity–on these other inputs to our happiness are not obvious. Growth and technological change enable greater heterogeneity in our attributes, making each of us more dissimilar and unique. While this freedom may be empowering, it could also make it harder for us to connect with each other.

A friend of mine recently mentioned he had been to a talk where the speaker discussed firms trying to promote more equitable outcomes for their customers. My friend was skeptical profit-maximizing firms would meaningfully deal with such issues barring state regulation. While I think this may be largely true, I think there are some counterbalancing reasons a firm might promote public goods a little more than initially expected.

First, firms fear future legal risk. The chance of regulation or legal proceedings may deter some bad behavior, even if regulations barring that behavior are not currently in place. For instance, firms may choose to not collect and use excessive low value data from customers if they anticipate possible future changes in the legal or regulatory landscape that would penalize holding such information. Sensitive customer information stored indefinitely may become difficult to manage because of future GDPR-type privacy regulations or may surface during legal proceedings and become very costly.

Second, altruistic behavior may be a marketing ploy. I expect most people view this as the dominant rationale for socially responsible firm behavior. If customers prefer to associate with brands that promote social good, then making charitable donations or engaging in other socially responsible behavior (and publicizing it) may in fact be profit-maximizing.

Third is that employees in a profit-maximizing firm may have interests in social welfare that compete with the firm’s interest in profit-maximization. If the firm does not have optimal control mechanisms to ensure their employees are working to maximize profits, the company’s direction may be drawn closer to the public good insofar as employees have altruistic preferences. This a classic example of the principal-agent problem, which explains how the firm may have difficulty assessing employees’ quality and effort. Inefficiencies in control may actually serve the public good if employees’ deviating behavior draws the firm to improved social outcomes instead of pure profit maximization. 

This form of altruism may in fact be a needed substitute to state intervention at moderating firm behavior in some settings. Can a stock of socially benevolent (but delinquent to the firm) employees promote public good when designing good regulations is difficult? It depends on context, but in cases with large information aggregation problems this may be one of the more effective mechanisms for promoting ethical behavior at firms. As Hayek described, a difficulty the government faces in regulating the economy is aggregating and processing information from massive networks of agents in a market, all of whom may have relevant private information they may be unable or unwilling to disclose. This problem may be even worse in cases where information and decision-making is particularly diffuse, like in firms with a large stock of highly-skilled labor. Labor that is scarce may be more difficult to adequately assess quality for; highly skilled workers may not have many co-workers who can precisely understand what exactly they do, how good it should be, and how long it should take. The firm’s difficulty in holding a worker accountable gives the worker more discretion in the direction of the work they do. Ideally, if they have non-profit maximizing preferences, they may use some of their freedom to maneuver the firm towards more ethical outcomes.

Cases where information in a firm is diffuse and employees have greater discretion are precisely when regulation may be the most challenging. If the firm itself cannot assess how well a worker is doing, how can a government acquire the needed information to design optimal policy? Employees are the ones actually executing the tasks of the company and most in-the-know about how the firm operates. Using that information they may nudge their respective functions in the firm towards better behavior. It is possible such altruistic behavior may in part compensate for poor conditions for regulation.

Of course, the opposite is also possible. Employees’ interest in their own rents may dominate whatever altruistic behavior they have, leading to worse outcomes than profit maximizing behavior. Less accountability may lead to higher likelihood of getting away with harassment or other seriously harmful actions. In firms with diffuse information networks, understanding the value of “good” employees on better firm behavior is an interesting and important empirical question.

Finally, employees may push firms to act in the interest of the public good for reasons other than just poor oversight. They may prefer to work on projects they care about, find interesting, and find fulfillment in. In that case allowing work that promotes public good or acquiescing to employee demands may be a method for acquiring and retaining talent. Talent acquisition concerns may also matter when employees face pressure from friends or family outside the firm who view its actions as unethical. Distaste for working for a firm perceived by the public as bad likely explains why JUUL, the e-cigarette maker, purportedly has to pay prospective employees a very significant premium to come work for them.

Still, it is hard to believe altruistic employee preferences or social pressure would be enough to optimally solve externality and public goods problems that we might like firms to address in an ideal setting. Better outcomes require policies and mechanisms to assess and encourage alignment of firm incentives and the public good.