Women's Liberation as a Financial Innovation

• Author: HOSNY ZOABI,DAVID WEISS,MOSHE HAZAN
• Date: 01 December 2019
• DOI: http://doi.org/10.1111/jofi.12829
• Published date: 01 December 2019

THE JOURNAL OF FINANCE •VOL. LXXIV, NO. 6 •DECEMBER 2019

Women’s Liberation as a Financial Innovation

MOSHE HAZAN, DAVID WEISS, and HOSNY ZOABI∗

ABSTRACT

In one of the greatest extensions of property rights in human history, common law

countries began giving rights to married women in the 1850s. Before this “women’s

liberation,” the doctrine of coverture strongly incentivized parents of daughters to hold

real estate, rather than financial assets such as money, stocks, or bonds. We exploit

the staggered nature of coverture’s demise across U.S. states to show that women’s

rights led to shifts in household portfolios, a positive shock to the supply of credit,

and a reallocation of labor toward nonagriculture and capital-intensive industries.

Investor protection thus deepened financial markets, aiding industrialization.

“It was now proposed that, for the first time in our history, the property of

one-half of the married people of this country should receive the protection

of the law. Up to this time the property of a wife had had no protection

from the law . . . ”

MP Russell Gurney, during the debate on the Married Women’s Property

Act of 1870. (Hansard, British House of Commons, April 14th, 1870).

PROPERTY RIGHTS ARE AT THE heart of capitalism’s ability to efficiently allocate

resources. In one of the greatest extensions of property rights in human history,

common law countries began giving rights to married women in the second

∗Moshe Hazan is with the Eitan Berglas School of Economics at Tel Aviv University and CEPR.

David Weiss is with the Eitan Berglas School of Economics at Tel Aviv University. Hosny Zoabi

is with The New Economic School. We thank the Editor, Stefan Nagel, two anonymous referees,

Nittai Bergman, Louis Cain, Doug Campbell, Alma Cohen, Matthias Doepke, Steven Durlauf,

Oded Galor, David Gilo, Gunes Gokman, Dror Goldberg, Jeremy Greenwood, Ron Harris, Fatih

Karahan, Patrick Kelly, Peter Koudijs, Joram Mayshar, Andrea Matranga, Stelios Michalopoulos,

Jean-Philippe Platteau, Hugh Rockoff, YonaRubinstein, Itay Saporta-Eksten, Mark Schankerman,

Analia Schlosser, Yannay Spitzer, Mich`

ele Tertilt, Guillaume Vandenbroucke, Yi Wen, Evgeny

Yakovlev, Yaniv Yedid-Levi, Joseph Zeira, Fabrizio Zilibotti, Ariell Zimran, and participants at

many conference and seminar presentations for helpful comments. We thank Efraim Benmelech,

Rick Geddes, Matt Jaremski, Kris Mitchener, and Robert Tamura for sharing data. We thank

Hal and Nancy Krent for helping understand court cases. Pavel Bacherikov, Alexey Khazanov,

and Yannay Shanan provided excellent research assistance. Zoabi acknowledges the Russian Sci-

ence Foundation, grant #18-18-00466 for supporting GIS-based analysis (Sections 5.1 and 5.3).

Hazan and Weiss gratefully acknowledge financial support from the Pinhas Sapir Center and the

Israel Science Foundation grant number 1705/16 for supporting non-GIS-based analyses. All of

the authors have read the Journal of Finance’s disclosure policy and have no conflicts of interest

to disclose.

DOI: 10.1111/jofi.12829

2915

2916 The Journal of FinanceR

half of the 19th century. Before this “women’s liberation,” married women were

subject to the laws of coverture.1Coverture had detailed regulations as to which

spouse had ownership and control over various aspects of property, both before

and after marriage, and strongly incentivized women to hold real estate, rather

than financial assets such as money, stocks, or bonds. This paper explores the

economic ramifications of coverture’s demise and the resulting expansion of

investor protection to women. In particular, we exploit the staggered nature

of coverture’s demise across the United States to show that these rights had a

large impact on household portfolios, credit markets, and labor allocations.

Under coverture, property was divided into two types. Moveable property

(also referred to as “personal property”), including money, stocks, bonds, fur-

niture, and livestock, became the husband’s property entirely upon marriage.

He could sell or give the property away, or even bequeath it to others. Real

assets, such as land and structures, were placed under the husband’s partial

control while remaining in the wife’s name. He could manage the assets as

he saw fit, including any income generated by the assets, but he could not

sell or bequeath the property without his wife’s consent.2After analyzing the

laws of coverture, Holcombe (1983) concludes that “[w]hatever the reasons for

the distinction between real and personal [moveable] property, the legal rules

applying to these categories of property were substantially different. The com-

mon law afforded married women considerable protection with respect to real

property. It afforded no protection for their personal property” (p. 20).

By differentially allocating property rights, coverture affected portfolio in-

centives not only for women, but also for parents wishing to bequeath or gift

assets to their daughters. Consider a father who wanted to bequeath his es-

tate to his daughter upon his death. He would face an incentive to hold his

wealth in real assets. Indeed, parents did bequeath to daughters in the United

States as primogeniture was abandoned after the War of Independence. The

default became to divide inheritances of both types of assets equally among

children, including girls (Shammas, Salmon, and Dahlin (1987), p. 67). There-

fore, our first prediction is that undoing coverture should cause portfolios to

shift toward moveable assets, such as financial assets, because removing legal

constraints allows households to purchase assets with higher returns or diver-

sify their portfolios.3Such a shift in portfolios toward moveable assets would

yield an increase in the supply of financial assets. Accordingly, our second

1Coverture was an inherent aspect of British common law,and as such applied both in England

and her colonies, including those that formed the United States, Canada, and Australia.

2See Blackstone (1896) for the laws of coverture. For a summary of the general responsibilities

that husbands and wives had to one another under coverture, see tables 1 and 2 of Basch (1982).

We discuss further details of the laws of coverture, as well as the origin of the differentiation of

the two types of property in Section Iof the Internet Appendix. The Internet Appendix is available

with the online version of this article on The Journal of Finance website.

3Baskerville (2008) studies the effects of women’s property rights in Canada and argues that,

after rights were granted, “[i]f one were to take away the very rich and obviously powerful, then

women’s activities and profiles in those areas [wealth holdings/portfolio choices] were often undis-

tinguishable from those of most of their male counterparts” (Baskerville (2008), p. 237).

Women’s Liberation as a Financial Innovation 2917

prediction is that after rights are granted, bank deposits—and loans—should

increase, whereas interest rates should decrease. An increase in the supply of

loanable funds would facilitate industrialization, as capital would be cheaper

and thus entrepreneurs could invest more readily. Greater industrialization

yields a sectoral reallocation of workers. Accordingly, our third prediction is

that coverture’s demise should lead to a shift in the labor force away from

agriculture. Moreover, within the nonagricultural sector, cheaper capital is ex-

pected to boost investment in industries that are more capital-intensive. Thus,

our fourth and final prediction is that rights should lead to a relative increase

in employment in capital-intensive industries.

Using the staggered nature of coverture’s demise across U.S. states, we

conduct four sets of empirical exercises to test these four predictions. Mas-

sachusetts was the first state to grant married women property rights in 1846.

By 1920, all but four states had followed suit. Geddes and Lueck (2002) argue

that it may not be fair to call the post-1920 era true coverture, as the 19th

Amendment (passed in 1920) granted women the right to vote. This may well

have affected the de facto implementation of coverture. Accordingly, we use

1850 to 1920 as our sample period whenever the data permit.

Endogeneity of women’s rights, in particular, omitted variable bias and re-

verse causality, is a natural concern. In Section III, we discuss these issues

using the historical record of British parliamentary debate, the academic lit-

erature on women’s rights, and empirical evidence from the United States. We

argue that our exercises can be viewed as capturing the causal impact of grant-

ing women rights on economic outcomes, especially in light of the border level

analyses described below.

Our first exercise uses population census data from 1860 and 1870, the only

two census years for which data on portfolios by asset class are available. We

show that households in states that granted rights during this decade increased

their holdings of moveable property as a fraction of their portfolios relative

to households in states that did not alter women’s economic status. Combs

(2005) uses a sample of British shopkeepers’ wives to provide evidence on how

property rights affected portfolio holdings in England. In contrast, we estimate

our regressions that both the whole United States as well as a sample including

state border pairs between states granting rights between 1860 and 1870 and

their neighbors. This latter methodology is in the spirit of Holmes (1998)but

uses a “difference-in-discontinuities” approach, as in Grembi, Nannicini, and

Troiano (2016), which allows us to control for geographic variation in economic

conditions. Our estimates suggest that the fraction of portfolios dedicated to

moveables increased by 1.0 to 6.3 percentage points. In addition, the number

of households holding any moveables increased by 1.3 to 7.4 percentage points

after rights were given, whereas the number of households holding any real

assets decreased by 1.7 to 6.9 percentage points.4We control for a variety of

4The shift from real to moveable assets may have had significant economic consequences beyond

sectoral reallocation. Ferrie (2003) shows that higher holdings of moveable assets during this

time period were associated with lower mortality rates, while there was no relationship between