This dissertation explores the role of informational frictions in macroeconomics and highlights how these frictions influence micro-level decisions which, when aggregated, result in more volatile macroeconomic fluctuations.

Chapter 1 explores the role of costly information and pricing complementarities in explaining persistent effects of monetary policy and behavior of prices at the micro level. In a setting where firms plan on when to acquire costly information, strategic complementarities in pricing generates planning complementarities. This results in a sluggish response of prices to monetary policy shocks. The calibrated model matches frequent and large price changes along with substantial non-neutralities. The chapter analyzes the effectiveness of monetary policy in the US since the 1970's and finds that it was relatively less effective in the 1970's compared to the subsequent decades.

Chapter 2 explores the role of dispersed opinions about economic conditions in reinforcing economic fluctuations and the role of policy to curtail these fluctuations. Output fluctuations arising from optimism and pessimism are often believed to be inconsistent with rational expectations. I show that dispersed information together with strategic complementarity, can give rise to endogenous cycles of pessimism and optimism which amplify these fluctuations. In the model, agents try to infer both true fundamentals and what others perceive the fundamentals of the economy to be, from both private signals and common prices. More precisely, agents "Forecast the forecasts of others". Correlated forecast errors mimic pessimism and optimism. These endogenously generated correlated forecast errors interact with production and investment decisions to cause volatile swings in both current and future output. Three key results emerge. First, an economy with dispersed information features amplified output fluctuations relative to the full information economy. Importantly, prices reinforce sentiments which in turn generate more volatile and persistent fluctuations. Second, in an otherwise neoclassical economy, dispersed information implies that the perception of aggregate demand matters in the output decision of firms. Finally, these fluctuations are inefficient and aggregate demand management through pro-cyclical payroll taxes or counter-cyclical sales subsidies can be reduce volatility and improve welfare without the need for the policy maker to have an informational advantage over the private sector.