Dennis Gärtner

We model retail-price recommendations (RPRs) as a communication device in vertical supply relations with private manufacturer information on production costs and consumer demand. With static trade, RPRs are irrelevant, and the equilibrium outcome is inefficient. With repeated trade, RPRs can become part of a relational contract, communicating private information from manufacturer to retailer that is indispensable for maximizing joint surplus. We show that this contract is self-enforcing if the retailer's profit is independent of production costs and punishment strategies are chosen appropriately. The predictions of our analysis are consistent with the available empirical evidence.

When labor incomes approach subsistence levels, the labor supply curve slopes outward, because the fight for survival mandates households to look for longer work hours in response to falling wage rates. We explore conditions under which near-subsistence scenarios may imply wage traps, labor market failures that can be the cause of undernourishment, illiteracy, and child labor. After stating general conditions under which wage traps occur, we look at specific production functions typically employed in quantitative analyses of growth and development. We find that standard Cobb-Douglas production functions do not permit wage traps, whereas CES functions do. Beyond that it turns out that when subsistence requirements increase with work hours, and when work effort rises with the wage rate, up to the efficiency-wage threshold, wage traps become more likely. Measures such as bans on child labor, implementation of minimum wage laws, or the establishment of labor unions may quite effectively improve conditions in wage-trapped labor markets.

This article investigates the design of incentives in a dynamic adverse selection Framework where agents' production technologies display learning effects and agents' learning rates are private knowledge. In a simple two-period model with full commitment available to the principal we show that whether learning effects are over- or underexploited crucially depends on whether more effcient agents also learn faster (so costs diverge through learning effects) or whether it is the less effcient agents who learn faster (so costs converge). We further show that an overexploitation of learning effects can occur also if the full-commitment assumption is relaxed.

This paper investigates the merger wave hypothesis for the US and the UK employing a Markov regime-switching model. Using quarterly data covering the last 30 years, for the US, we identify the beginning of a merger wave in the mid 1990s but not the much-discussed 1980s merger wave. We argue that the latter finding can be ascribed to the refined methods of inference offered by the Gibbs sampling approach. As opposed to the US, mergers in the UK exhibit multiple waves, with activity surging in the early 1970s and the late 1980s.

We consider a setting in which two potential merger partners each possess private information pertaining both to the profitability of the merged entity and to stand-alone profits, and we investigate the extent to which this private information makes ex-post regret an unavoidable phenomenon in merger negotiations. To this end, we consider ex-post incentive compatible mechanisms, which use both players' reports to determine whether or not a merger will take place and what each player will earn in each case. When the outside option of at least one player is known, the efficient merger decision can be implemented by such a mechanism under plausible budget-balance requirements. When neither outside option is known, we Show that the potential for regret-free implementation is much more limited, unless the budget balance condition is relaxed to permit money-burning in the case of false reports.

This paper examines the effects of introducing competition into monopolized network industries on prices and infrastructure quality. Analyzing a model with reduced-form demand, we first show that deregulating an integrated monopoly cannot simultaneously decrease the retail price and increase infrastructure quality. Second, we derive conditions under which reducing both retail price and infrastructure quality relative to the integrated monopoly outcome increases welfare. Third, we argue that restructuring and setting very low access charges may yield welfare losses as infrastructure investment is undermined. We provide an extensive analysis of the linear demand model and discuss policy implications.

This paper investigates how leniency programs induce collusive offenders to self report in a dynamic setting, where the risk of independent detection evolves stochastically over time. We show how uncertainty about the future can push firms into preemptive application, and how these preemptive incentives can unravel to the point where firms apply long before the risk of detection is in any way imminent. For policy, the analysis suggests that a relatively harsh treatment of latecomers may be more important than offering high absolute reductions or even rewards, and that leniency should be available to firms who are already under investigation.