We study the effects of advertising disclosure regulations in social media markets. Using data from a large sample of Instagram influencers in Germany and Spain and a difference-in-differences approach, we empirically evaluate the effects of German strengthening of disclosure regulations on post content and follower engagement. We measure whether posts include suggested disclosure terms and use text-based approaches (keywords, machine learning) to assess whether a post is sponsored. We show substantial adoption of disclosure but also a 12% increase in sponsored content and an increase in the share of undisclosed-sponsored content consumers are exposed to. We also find reductions in engagement, suggesting that followers were likely negatively affected.

Motivated by markets for "expertise," we study a bandit model where a principal chooses between a safe and risky arm. A strategic agent controls the risky arm and privately knows whether its type is high or low. Irrespective of type, the agent wants to maximize duration of experimentation with the risky arm. However, only the high type arm can generate value for the principal. Our main insight is that reputational incentives can be exceedingly strong unless both players coordinate on maximally inefficient strategies on path. We discuss implications for online content markets, term limits for politicians, and experts in organizations.

Consumers rely on intermediaries ("influencers") such as social media recommendations to provide information about products. The advice may be mixed with endorsement in a way that is unobservable to the follower, creating a trade-off for influencers between the best advice and the most revenue. This article models the dynamic relationship between an influencer and a follower. The relationship evolves between periods of less and more revenue. The model can provide insight into policies such as the Federal Trade Commission's mandatory disclosure rules. An opt-in policy may be superior: it deregulates influencers who are reaping the rewards of past good advice.

Targeted economic development subsidies do not work as advertised. In fact, the balance of economic theory and empirical evidence suggests that they are more likely to undermine development than to enhance it. Yet policymakers face strong incentives to continue to offer subsidies. Because subsidies are economically costly but politically valuable, they create a situation similar to a prisoner’s dilemma. An interstate compact offers a solution by changing the political payoffs. Importantly, interstate compacts enable policymakers to credibly commit to ending what many already see as a race to the bottom. Ending the mutually destructive subsidy war would allow state and local governments to repurpose up to $95 billion annually to tax relief and other projects with better payoffs.

We collect data from fifty top Instagram influencers in Germany and Spain from 2014 to 2019. Germany experienced changes in disclosure regulation for social media sponsorship during the sample period. Using a difference-in-difference approach, we study the impact of the the rules on the content of posts and the nature of interaction of followers with the posts. On the content side, we measure whether posts include suggested disclosure terms and show variable but substantial adoption of disclosure. We use an approach based on a fixed list of words associated with sponsorship (i.e. links, mentions of brands, use of words like "sale") as well as natural language processing to assess the likelihood that a post is sponsored. We show that sponsored content use may have increased after changes in disclosure and that followers may have been negatively affected. On the other hand, there is evidence that consumers' reaction to sponsored posts, measured by likes, may be quite different under stricter disclosure rules, suggesting that the rules could have a substantial impact on information transmission.

Motivated by reputation management in a variety of different markets for ``expertise'' (such as online content providers and experts in organizations), we develop a novel repeated-game framework in which a principal screens a strategic agent whose type determines the rate at which he privately receives payoff relevant information. The stage game is a bandit setting, where the principal chooses whether or not to experiment with a risky arm which is controlled by an agent who privately knows its type. Irrespective of type, the agent strategically chooses output from the arm to maximize the duration of experimentation. Experimentation is only potentially valuable to the principal if the arm is of the high type. Our main insight is that reputational incentives can be exceedingly strong: the agent makes inefficient output choices in all equilibria (subject to a mild refinement) and that this can result in market breakdown even when the uncertainty about the agent's type is arbitrarily small. We show that (one-sided) transfers do not prevent this inefficiency and we suggest alternate ways to improve the functioning of these markets.

The past decade has witnessed a resurgence in innovation awards, in particular of grand innovation prizes (GIPs) which are rewards to innovators developing technologies reaching performance goals and requiring breakthrough solutions. GIPs typically do not preclude the winner also obtaining patent rights. This is in stark contrast with mainstream economics of innovation theories where prizes and patents are substitute ways to generate revenue and encourage innovation. Building on the management of innovation literature which stresses the difficulty to specify ex-ante all the technical features of the winning technologies, we develop a model in which innovative effort is multi-dimensional and only a subset of innovation tasks can be measured and contracted upon. We show that in this environment patent rights and cash rewards are complements, and that GIPs are often preferable to patent races or prizes requiring technologies to be placed in the public domain. Moreover, our model uncovers a tendency for patent races to encourage speed of discovery over quality of innovation, which can be corrected by GIPs. We explore robustness to endogenous entry, costly public funds, and incomplete information by GIP organizers on the surplus created by the technology.

This article introduces an economic model of dynamic capabilities. The model is intended to bridge the gap between the strategic management literature on dynamic capabilities and the economics literature on the sources of productivity differentials. In the model, dynamic capabilities are an advantage in generating innovations and allow firms to enter new submarkets. We use the model to interpret three types of dynamic capabilities previously identified in the literature: sensing, seizing and transforming. We show that the model has non-trivial predictions for observables that might aid in empirically identifying dynamic capabilities. When firms must invest in order to acquire the dynamic capability to transform, in a dynamic equilibrium, higher levels of innovation may not be associated with possessing the dynamic capability: firms without the dynamic capability can invest more in innovation in order to gain it. This suggests that using innovation investment levels as an intermediate outcome to measure dynamic capabilities may incorrectly identify firms without the dynamic capability as ones with it.

We study how best to reward innovators whose work builds on earlier innovations. Incentives to innovate are obtained by offering innovators the opportunity to profit from their innovations. Since innovations compete, awarding rights to one innovator reduces the value of the rights to prior innovators. We show that the optimal allocation involves shared rights, where more than one innovator is promised a share of profits from a given innovation. We interpret such allocations in three ways: as patents that infringe on prior art, as licensing through an optimally designed ever-growing patent pool, and as randomization through litigation. We contrast the rate of technological progress under the optimal allocation with the outcome if sharing is prohibitively costly, and therefore must be avoided. Avoiding sharing initially slows progress, and leads to a more variable rate of technological progress.

We analyze a model of industry evolution where the number of active submarkets is endogenously determined by pioneering innovation from incumbents and entrants. Incumbent pioneers enjoy an advantage of additional pioneering innovation via a dynamic capability that takes the form of an improved technology for innovation in young submarkets. Entrants are motivated in part by a desire to acquire the dynamic capability. We show that dynamic capabilities increase total innovation, but whether the capability confers an advantage in terms of marginal or average cost is important in determining how the impact of dynamic capabilities is distributed across incumbent and entrant innovation rates. We complement the existing literature—that focuses on exogenous arrival of submarkets or the steady state of a model with constant submarkets—by describing how competition, free entry, and the dynamic capability of incumbents drive the evolution of an industry. The shift from immature to mature submarkets can lead to a shakeout in firm numbers, and it eventually leads to a reduction in total dynamic capabilities in an industry. This paper was accepted by Bruno Cassiman, business strategy.

This paper considers a moral hazard problem where a principal contracts with one or more agents to produce a design. The design refers to something whose value is subjective, may or may not be well understood by the principal, and can be returned if the principal deems it to have low value, which is sometimes called right of refusal. Correlation of the principal's signal with the his true value plays a role, in contrast to standard principal agent problem; experience goods are di erent from credence goods. The principal's ability to forecast value corresponds to a notion of taste for the principal that distinguishes taste from judgment. This measure of informativeness of the signal matters for many features of the contract design: the total cost, the number of agents contracted with, and the choice of agent quality. This measure of informativeness is related to subjectivity, and shows that privateness of the signal, which has been used synonymously with subjectivity in the literature, does not fully describe subjectivity. For examples where the subjectivity comes through the taste rather than lack of judgment, uncertainty in the outcome may make the incentive contract less costly, in contrast to both the canonical model of moral hazard and the case where signals are private but not necessarily truly subjective. ∗University of Toronto. Thanks to Heski Bar-Isaac, Jim Campbell, Rahul Deb, Marcin Peski, and the Toronto Theory and IO Bag Lunches for comments.

We study how to reward innovators who build on one another. Rewards come in the form of patents. Because patent rights are scarce, the optimal allocation involves sharing: More than one innovator's patent is in force at a given time. We interpret such allocations as patents that infringe one another as licensing through an ever growing patent pool and as randomization through litigation. We contrast the rate of technological progress under the optimal allocation with the outcome if sharing is prohibitively costly. Avoiding sharing initially slows progress and leads to a more variable rate of technological progress.

Patents are a useful but imperfect reward for innovation. In sectors like pharmaceuticals, where monopoly distortions seem particularly severe, there is growing international political pressure to identify alternatives to patents that could lower prices. Innovation prizes and other non-patent rewards are becoming more prevalent in government's innovation policy, and are also widely implemented by private philanthropists. In this paper we describe situations in which a patent buyout is effective, using information from market outcomes as a guide to the payment amount. We allow for the fact that sales may be manipulable by the innovator in search of the buyout payment, and show that in a wide variety of cases the optimal policy still involves some form of patent buyout. The buyout uses two key pieces of information: market outcomes observed during the patent's life, and the competitive outcome after the patent is bought out. We show that such dynamic market information can be effective at determining both marginal and total willingness to pay of consumers in many important cases, and therefore can generate the right innovation incentives.

An important type of innovation is one that pioneers a new submarket. Klepper and Thompson (2006) and Sutton (1998) show that innovation driven by the scope of the market can explain a variety of empirical facts. We introduce a model where innovators must decide whether to pioneer a new submarket or compete in an existing one. Unlike the prior research on pionnering submarkets and industry evolution, we endogenize the existince of submarkets to pioneer and show how the model generates an equilibrium long run scope of the market. We show that the model can explain some of the important existing facts about the product life cycle. We investigate the product cycle data and show other important features that are consistent with the model. Finally, we show that the model has implications for policy directed at particular kinds of innovations.

We study an environment where a duopoly develops innovations that build on one another. The quality of innovations is private information, so optimal rewards take the form of rights to produce the resulting products. There is a tradeo between encouraging one rm to work on its innovations by granting it promised rights, and the fact that those rights deteriorate the rights of its competitors. We study constrained e cient allocations and show that they result in nearpermanent monopolization: eventually one rm is promised nearly everything, and the competitor is almost completely ignored. This occurs because backloading rewards is an e cient incentive device. We interpret our results in three ways. First, if one thinks of our allocations as o ering policy guidance, then protection is state-dependent as in [1], generating heterogeneity in patent protection in the absence of heterogeneity in innovation opportunities. We show how this protection can be simply implemented by allowing rms to buy additional protection. Second, if the rms are able to contract ex ante, our allocations can be interpreted as a patent pooling arrangement between the rms, where one rm comes to dominate pool membership. Finally, one can interpret the optimal evolution of the duopoly as competition for the market, where static competition is not the key form ∗We thank participants at numerous seminars and conferences for helpful comments. This paper was previously titled Optimal Patent Policy with Recurrent Innovators. †UCLA. ‡University of Toronto.

Technological progress is typically a result of trial-and-error research by competing firms. While some research paths lead to the innovation sought, others result in dead ends. Because firms benefit from their competitors working in the wrong direction, they do not reveal their dead-end findings. Time and resources are wasted on projects that other firms have already found to be dead ends. Consequently, technological progress is slowed down, and the society benefits from innovations with delay, if ever. To study this prevalent problem, we build a tractable two-arm bandit model with two competing firms. The risky arm could potentially lead to a dead end and the safe arm introduces further competition to make firms keep their dead-end findings private. We characterize the equilibrium in this decentralized environment and show that the equilibrium necessarily entails significant efficiency losses due to wasteful dead-end replication and a flight to safety – an early abandonment of the risky project. Finally, we design a dynamic mechanism where firms are incentivized to disclose their actions and share their private information in a timely manner. This mechanism restores efficiency and suggests a direction for welfare improvement. Ufuk Akcigit Department of Economics University of Pennsylvania 3718 Locust Walk, #445 Philadelphia, PA 19104 and NBER uakcigit@econ.upenn.edu Qingmin Liu Columbia University 1022 International Affairs Building Mail Code 3308 420 West 118th Street New York, NY 10027 qingmin@econ.upenn.edu

may, on their own, generate monotone matching predictions in the absence of complementarities or anti-complementarities in production technology. We also derive sufficient conditions on the primitives of the model leading to the optimality of positive and negative matching of team members.