Abstract
This paper examines the responsiveness of investment to Tobin's q using data from Real Estate Investment Trusts (REITs). We take advantage of high-quality estimates of the market value of a firm's assets instead of the more-commonly used book value of assets to estimate a more accurate value of q. We have three main results. First, while REITs have institutional features that mitigate many of the complications faced by previous studies, we still find little relationship between REIT investment and a traditional accounting-based measure of q. Yet our appraisal-based q generates elasticities of 43 percent, as much as 8 times larger than in previous studies. Second, when we decompose measurement error over time or remove a few large outliers, our estimates of the elasticity of investment with respect to q grow as high as 107%. Third, we show that other factors beyond Tobin's original model are still quite important. With our improved measure of q, cash flow still impacts investment, even the portion of cash flow that REITs are required by IRS rules to payout to shareholders as dividends and cannot be spent on investment. Also, we show that the relationship between our improved q and investment is nonlinear; with little impact of q on investment when q < 1, but very large elasticities when q exceeds 1.