This paper examines the relationship between real balances (narrow or broad) and industrial output in Bangladesh with seasonally-adjusted monthly data for the period 1973:01-2003:06. The results of the ADF and the KPSS tests suggest that although real narrow money balances have a unit root, real broad money balances and industrial output do not seem to have a unit root. Therefore, the paper applies the bounds testing approach of Pesaran et al. (2001) to investigate the long-run relationship between real balances and industrial output, given that this approach remains valid `irrespective of whether the underlying regressors are purely I(0), purely I(1) or mutually cointegrated'. The empirical results suggest a long-run relationship between real balances and industrial output, while neither of them can be treated as a `long-run forcing variable' for explaining the other. In the associated short-run model, the coefficient of the error-correction term bears the expected negative sign and is significant, which confirms the presence of a long-run relationship between real balances and industrial output. The forecasting ability of the error-correction model is satisfactory, especially with the broad definition of money. The overall results are consistent with the view that in an underdeveloped financial system, real money balances (especially broad money balances) should be considered a complementary, if not a primary, factor of production.
In an early contribution to the monetary growth theory, Tobin (1965) suggested a link of faster money supply growth with higher capital stock and output per person in the steady state. His argument was that in a monetary economy, a higher rate of inflation (caused by a higher rate of growth of the money supply) would lead savers to shift their portfolios in favor of capital (raising capital intensity) given that money pays no return but loses value with the rise in inflation. In monetary growth literature this is known as the Tobin effect (Orphanides and Solow, 1990). While the Tobin model establishes the argument against the super-neutrality of money, the classic defence of super-neutrality of money has come from Sidrauski (1967). He has shown that the sizes of the steady-state capital stock and real consumption are independent of inflation or the rate of money supply growth, implying that money is indeed super-neutral. Importantly, although Tobin's view on the effect of inflation on per-capita income appears plausible, it has a disturbing implication. For example, in Tobin's model, inflation, which acts as the opportunity cost of holding money, lowers money balances and raises capital intensity. This implies that if inflation rises on a sustained basis, this will ultimately force real balances to zero. This would be equivalent to the regression of a monetary economy to a barter economy. A barter economy is certainly less efficient than a monetary economy. Tobin has acknowledged this phenomenon and the welfare loss due to inflation (Haliassos and Tobin, 1990).
In Sidrauski's model, the marginal product of capital depends on the level of capital stock. As his model shows, real balances are inversely related to the rate of money creation in the steady-state, and the super-neutrality results cannot be sustained if real balances enter in the production function. Therefore, an important variation of his model is the inclusion of money as a factor of production. It is assumed that firms hold real balances to facilitate production, so that money acts as a complementary factor to capital. This can be expressed in a simple neoclassical production function: y = y(k,m), where y, k and m are the output, capital and real balances in per-capita terms. Note that implicit in the two-factor production function y = y(k) is the assumption that real balances do not affect output, or that the marginal productivity of real balances is zero. Many monetary economists consider this a restrictive assumption. The argument is that although it is difficult to treat real balances as a productive service in the same way as labor and capital are treated in the production function, it is an oversimplification to ignore the role of real balances in the production process. This issue has lately become prominent in the literature on monetary policy. |
