Labor Market Policies in a Deep Recession: Lessons from Hoover's Policies during the U.S. Great Depression

• Published date: 01 February 2022
• Author: JORDAN ROULLEAU‐PASDELOUP,ANASTASIA ZHUTOVA
• Date: 01 February 2022
• DOI: http://doi.org/10.1111/jmcb.12798

DOI: 10.1111/jmcb.12798

JORDAN ROULLEAU-PASDELOUP

ANASTASIA ZHUTOVA

Labor Market Policies in a Deep Recession:

Lessons from Hoover’s Policies during the

U.S. Great Depression

What are the effects of labor market policies when the economy is sliding

into a deep recession? Weshow that public announcements asking rms not

to cut wages under H. Hoover during 1929–33 postponed entering the zero

lower bound episode and reduced its duration. We develop and estimate a

medium scale New Keynesian model to measure the effect of Hooverpoli-

cies during the Great Depression and we nd evidence that without such

polices the U.S. economy would have ended up in a liquidity trap 3 years

before it actually did, suffering an evendeeper recession with a larger dea-

tion.

JEL codes: C11, E24, E31, E32, E44, E52, N12

Keywords: Zero lower bound, Deation, Great Depression

T     large literature studying the effect of

various policies once an economy enters a liquidity trap.1However, much less is

known about which policies can prevent (or at least postpone) the economy from

Wethank Yvan Becard, Kenza Benhima, Fabrizio Coricelli, Gauti Eggertsson, Céline Poilly, Alejandro

Justiniano, Etienne Lehmann, Antoine Lepetit, Jean-Baptiste Michau, Lee E. Ohanian, Stephanie Schmitt-

Grohé, Martin Uribe, and severalseminar participants for interesting comments. Weare especially grateful

to Jean-Olivier Hairault and Florin O. Bilbiie for their extensive comments. Part of the work on this pa-

per by Jordan Roulleau-Pasdeloup has been done at the Centre de Recherche en Économie et Statistique

(CREST) during his Ph.D. and he would like to thank them for their kind hospitality and support. The

views expressed herein are those of the authors and should not be interpreted as reecting those of the

Banque de France.

J R-Pis an Assistant Professor, Department of Economics, National Uni-

versity of Singapore (E-mail: jordan.roulleau@gmail.com). A ZisResearch Economist,

Banque de France (E-mail: Anastasia.ZHUTOVA@banque-france.fr).

Received June 9, 2020; and accepted in revised form October 14, 2020.

1. See, among many others, Eggertsson (2008), Christiano, Eichenbaum, and Rebelo (2011), Eggerts-

son (2011), Woodford (2011), and Eggertsson (2012).

Journal of Money, Credit and Banking, Vol. 54, No. 1 (February 2022)

© 2021 The Ohio State University

248 :MONEY,CREDIT AND BANKING

sliding into a liquidity trap in the rst place.2Given that monetary policy is usually

best described as a feedback rule from ination (or the price level) to interest rates,

any policy that can limit deation after a recessionary shock can potentially postpone

the occurrence of a zero lower bound (ZLB) episode. In this paper, we study one

example of such a policy during the U.S. Great Depression and assess quantitatively

whether it did postpone the occurrence of zero interest rates.

The policies that we study are public announcements made by H. Hoover during

1929–33, with the objective to encourage growth in real wages; see Ohanian (2009)

and Rose (2010). Essentially, Hoover encouraged rms not to decrease their nominal

wages in return for keeping union demands at bay. The main result of this paper is

to show that, without these policies, the U.S. economy would have experienced an

even larger deation during the early stages of the Great Depression. By promoting

high wages, Hoover policies had the effect to dampen the fall in real marginal cost

of rms. Under the premise that prices are set as a markup over real marginal cost,

these policies limit deation.

Within our framework, in a counterfactual without Hoover policies, a stronger de-

ation pushes nominal interest rates further down. In turn, hitting the ZLB generates

a lower output that reinforces the fall in ination. With this in mind, another contri-

bution of our paper is to show that regardless of any initial negative impact on output

through labor market adjustments, Hoover policies were overall benecial since they

postponed entering the ZLB and prevented the associated fall in GDP. The magni-

tude of this effect is ultimately a quantitative question, and in this paper we set out to

answer it by developing a medium-scale model that we take to the data.

Recent work on the subject such as Ohanian (2009) concludes that promoting high

wages during a recession will postpone the necessary adjustments, making the crisis

more severe.3This argument is based on an Real Business Cycle (RBC) model with

exible prices, in which the ZLB constraint does not enter the picture. In this paper,

we show that one gets a very different result when considering both sticky prices and

the ZLB constraint on the nominal interest rate.

Our paper is also related to a recent series of papers by Gauti B. Eggertsson.4While

we focus on what happens before hitting the ZLB, he focuses on the dynamics once

the economy is already at the ZLB.5In a simple New Keynesian model, he shows

that policies of cartelization like Roosevelt’s infamous New Deal are expansionary

since they decrease expected real interest rate and thus increase aggregate demand.

2. Benhabib, Schmitt-Grohe, and Uribe (2002) consider various monetary and scal policies to avoid

the liquidity trap equilibrium associated with Taylor rules.

3. See also Cole and Ohanian (2004), Ohanian (2009), and Cole and Ohanian (2013).

4. See Eggertsson and Pugsley (2006), Eggertsson (2008), and Eggertsson (2012).

5. In related and independent work, Glover (2019a) discusses the impact of increasing the minimum

wage in the United States just before the Great Recession and shows that it mitigated the decrease in in-

ation.

JORDAN ROULLEAU-PASDELOUP,AND ANASTASIA ZHUTOVA :249

Conversely, it can be shown that when the economy is in a liquidity trap, austerity

policies will be contractionary.6

In our counterfactual without Hoover’s policies, the economy ends up at the ZLB

earlier than it did. Recent contributions such as Cohen-Setton, Hausman, and Wieland

(2016), GarÃn, Lester, and Sims (2019), and Wieland (2019) show that the conven-

tional New Keynesian model makes predictions that are at odds with what can be

found in the data.7One then has to wonder whether the model that we use here is

subject to these limitations. We argue that this is actually not the case. In our model,

the Central Bank follows a price-level targeting rule. Since the seminal work of Eg-

gertsson and Woodford (2003), we knowthat a price-level targeting rule comes close

to replicating the Ramsey-optimal monetary policy under full commitment. In turn,

most of the puzzles that have been shown to plague the conventional New Keyne-

sian model at the ZLB vanish under this setup for monetary policy. For example,

Roulleau-Pasdeloup (2018) shows that government spending increases at the ZLB

do not become unreasonably large when monetary policy is optimal.

To gauge empirically the effects of Hoover policies, we develop a medium-scale

New Keynesian model with nancial frictions as in Christiano, Motto, and Rostagno

(2003), ,2014). To account for Hoover policies, we explicitly model union negotia-

tion. Unions negotiate on behalf of union members, whose desired wage is an exoge-

nous markup over the marginal rate of substitution between leisure and consumption.

We call this shock an aspiration wage shock and take it to represent the potential ef-

fects of Hoover’s policies. Our results do not hinge on the specic choice of the labor

market setup. We obtain similar results with a more standard way of modeling labor

supply shocks (àlaErceg, Henderson, and Levin 2000). However, for our analysis

we use a model that can provide a meaningful explanation of a decrease in employ-

ment for a given real wage consistent with the evidence provided in Ohanian (2009)

and Rose (2010).

In our setup the Federal Reserve targets the price level, an assumption that is made

in many studies of the same period and considered to approximate rather well the

Gold Standard regime.8We estimate a log-linear version of the model on quarterly

data using nine macro-economic time series for the period of 1922:Q3–1932:Q3 with

Bayesian methods. A series of negative aggregate demand shocks generates dea-

tion. As a response, the Central Bank will lower its interest rate as long as it is not

constrained by the ZLB. It turns out that the estimation results unveils an important

role to the aspiration wage shock during the period 1929:Q4–1932:Q3. This is pre-

cisely the time when President Hoover gave his two speeches in front of the major

6. See the example of the Mistake of 1937 in Eggertsson and Pugsley (2006).

7. The main counterfactual prediction of this model is that positive supply shocks (like a temporary

increase in technology) have negative effectson output at the ZLB.

8. We will discuss this assumption in more details later.