Observability and Social Interactions in Consumption

This paper examines how social context affects consumption responses to transitory income shocks. Using a randomized survey experiment embedded in the Italian Survey of Consumer Expectations, we vary whether an unexpected €1,000 employer-specific bonus is observable to coworkers and whether it is received only by the respondent or by all coworkers. The design separately identifies the effect of observability for an idiosyncratic bonus and the incremental effect of making a publicly observable bonus common within the workplace, a well-defined social context. Making an idiosyncratic bonus publicly observable reduces the MPC by about 3 percentage points relative to a private idiosyncratic bonus. Conditional on observability, making the bonus common among coworkers raises the MPC by about 6 percentage points. Interpreted through a linear social-interactions model, the latter effect implies a socialinteraction parameter of approximately 0.16 and a social multiplier of about 1.2. These findings suggest that aggregate consumption responses to fiscal policy depend not only on individual behavior but also on the social context in which income changes occur.