A Jobs Guarantee alone cannot be an effective automatic stabilizer

While the MMT literature has not explicitly discussed the issue of time consistency, MMT economists have advocated for the use of automatic stabilizers as a means of overcoming general political economy concerns (Fullwiler, Grey, and Tankus 2019). Typically, their recommendations include procyclical tax policies and countercyclical spending, such as the JG (Wray 2019b, p. 17). The JG is intended to complement these programs rather than replace them entirely. Examples include traditional taxation policies, as well as "no longer indexing tax brackets or indexing them to an inflation target instead and introducing more tax brackets so that as incomes rise faster than the inflation target a higher percentage of income is progressively taxed" (Fullwiler, Grey, and Tankus 2019). For example, as incomes decline with decreased private sector demand, tax receipts fall. Note that tax bracket indexation may be counterproductive because changes in the price level may come from either a supply shock or a demand shock. Indeed, under this policy, the automatic stabilizer would be contractionary when an adverse supply shock occurs, which is precisely the opposite of the appropriate policy response. Additionally, rather than allowing a fiscal authority to attempt to change taxes in realtime, Fullwiler, Grey, and Tankus (2019) suggest that "varying tax rates and other inflation offsets should be included in the budgeting process from the outset." This may have some merit, but including "inflation offsets" at the beginning of the period does not categorically evade the problem of discretion if the rule is not independent of the fiscal decisionmaking process, which by design occurs at the beginning of the period. Moreover, the authors take the position that "we are not against one or more agencies being given additional tools to collectively manage demand on a discretionary basis" (Fullwiler, Grey, and Tankus 2019) yet we fail to see how this eliminates fiscal fine-tuning. While we have no objection to further progressivity on stabilization grounds and noting that tax policies are already procyclical, we do observe several problems with relying on procyclical tax policies as a primary stabilizer. First, there is a prime mover problem. Suppose there is a demand shock primarily due to loose fiscal policy. That is, automatic stabilizers are not designed sufficiently well such that inflation could be throttled at its source. Presumably, this is not impossible. Then we are back to our model: discretionary policy is necessary to reduce inflation any further than the automatic stabilizers have accomplished. Second and relatedly, it is improbable that, absent an enormous change in the structure of taxation, automatic tax policy would be sufficient to reduce the BER significantly absent discretionary intervention. Historically, discretionary policy has made up 50% of the policy response to demand shocks (Sheiner and Ng 2019). Indeed, Auer23 bach and Feenberg (2000) and Cohen and Follette (2000) find that the impact of automatic stabilizers, while apparent, is overall relatively modest, indicating that MMT may have a more substantial overhaul of the tax system in mind. Otherwise, the policies advocated would be insufficient. Moreover, the automatic stabilizers would have to be designed to automatically move the BER to the NAIBER, something which would require current knowledge of the NAIBER but perfect foresight of the structure of the economy. The reasoning is simple: the structure of the economy necessarily changes over time, which implies that automatic stabilizers cannot be designed ex ante such that inflation is perfectly addressed in each case. Hence some form of discretion is required. This does not necessarily mean tax policy, but it does mean that at some point the fiscal authority has to make a choice.12 At some point, there must be discretion. Third, and perhaps most importantly, it is impossible for these automatic stabilizers to discriminate between supply shocks and demand shocks (Blanchard 2000). In an ideal world, stabilization policies would minimize deviations from potential output but not react to changes in potential output. This is not the case with automatic stabilizers (like tax policy) intended to react to changes in output, but which cannot adequately discriminate between changes in potential output and deviations from potential output. Moreover, as noted by Blanchard and Summers (2020), output shocks tend to be more persistent than unemployment shocks and hence may be permanent changes to potential output rather than deviations from potential output. Consequently, automatic stabilizers triggered by output changes—like tax policy—are not recommended. Fourth, Blanchard and Summers (2020) advocate for semi-automatic stabilizers triggered by unemployment rather than output. Semi-automatic stabilizers are fiscal rules— 12. Moreover, the design of automatic stabilizers would not be immune from the Lucas Critique (Lucas 1976). generally, tax or spending measures—"triggered by the crossing of some statistical threshold, be it a low output growth rate or a high unemployment rate" (Blanchard and Summers 2020, p. 125). An existing example is the extension of unemployment benefits during a recession. In a JG world, this is not possible. Changes in the BER follow changes in the stance of fiscal policy rather than vice versa, so that the structure of JG employment could not serve as a useful trigger for an automatic stabilizer. Thus, absent a major innovation in stabilization policy, MMT policymakers would conduct automatic stabilizers (or semi-automatic stabilizers) based on output triggers, a suboptimal policy. Even taking into account automatic stabilizers like a progressive income tax structure, the fiscal authority would still, on net, prefer fewer taxes than are required to move the BER to the NAIBER. Moreover, the only way to avoid this outcome, which would be to eliminate discretion, would require perfectly designing a system of automatic stabilizers. Then, the future path of booms and busts is exactly offset, and the BER equals the NAIBER consistently, something which we have already argued is impossible. Finally, if the fiscal authority does tie its own hands and leaves demand management completely to an imperfect system of automatic stabilizers, there would inevitably be accelerating or decelerating inflation. The reason is simple: if fiscal policy is necessary to move the BER to the NAIBER, but is incapable of doing so, then the BER will be at a level such that inflation is accelerating or decelerating.