Economic Growth Second Edition
The Importance of the Transversality Condition
Download 0.79 Mb. Pdf ko'rish
|
BarroSalaIMartin2004Chap1-2
2.6.2
The Importance of the Transversality Condition It is important to emphasize the role of the transversality condition in the determination of the unique equilibrium. To make this point, we consider an unrealistic variant of the Ramsey model in which everyone knows that the world will end at some known date T > 0. The utility function in equation (2.1) then becomes U = T 0 u[c (t)] · e nt · e −ρt dt and the non-Ponzi condition is a (T ) · exp − T 0 [r (v) − n] dv ≥ 0 The budget constraint is still given by equation (2.3). Since the only difference between this problem and that of the previous sections is the terminal date, the only optimization condition that changes is the transversality condition, which is now a (T ) · exp − T 0 [r (v) − n] dv = 0 Since the exponential term cannot be zero in finite time, this condition implies that the assets left at the end of the planning horizon equal zero: a (T ) = 0 (2.32) In other words, since the shadow value of assets at time T is positive, households will optimally choose to leave no assets when they “die.” The behavior of firms is the same as before, and equilibrium in the asset markets again requires a (t) = k(t). Therefore, the general-equilibrium conditions are still given by equa- tions (2.24) and (2.25), and the loci for ˙ˆk = 0 and ˙ˆc = 0 are the same as those shown 22. Similar results apply if the economy begins with ˆk (0) > ˆk ∗ in figure 2.1. The only complication here is that, if investment is irreversible, the constraint ˆc ≤ f (ˆk) may be binding in this region. See the discussion in appendix 2B (section 2.9). Growth Models with Consumer Optimization 105 in figure 2.1. The arrows representing the dynamics of the system are also the same as before. Since a (t) = k(t), the transversality condition from equation (2.32) can be written as ˆk (T ) = 0 (2.33) From the perspective of figure 2.1, this new transversality condition requires the initial choice of ˆc (0) to be such that the capital stock equals zero at time T . In other words, optimality now requires the economy to land on the vertical axis at exactly time T . The implication is that the stable arm is no longer the equilibrium, because it is does not lead the economy toward zero capital at time T . The same is true for any initial choice of consumption below the stable arm. The new equilibrium, therefore, features an initial value ˆc (0) that lies above the stable arm. It is possible that ˆc and ˆk would both rise for awhile. In fact, if T is large, the transition path would initially be close to, but slightly above, the stable arm shown in figure 2.1. However, the economy eventually crosses the ˙ˆk = 0 schedule. Thereafter, ˆc and ˆk fall, and the economy ends up with zero capital at time T . We see, therefore, that the same system of differential equations involves one equilibrium (the stable arm) or another (the path that ends up on the vertical axis at T ) depending solely on the transversality condition. Download 0.79 Mb. Do'stlaringiz bilan baham: |
Ma'lumotlar bazasi mualliflik huquqi bilan himoyalangan ©fayllar.org 2025
ma'muriyatiga murojaat qiling
ma'muriyatiga murojaat qiling