comments on matsa and millerusers.nber.org/~jwolfers/papers/comments/comments on...justin wolfers,...

Comments on David Matsa and Amalia Miller

“A Female Style in Corporate Leadership?

Evidence from Quotas”

Justin Wolfers

Wharton School, University of Pennsylvania

Brookings, CEPR, CESifo, IZA and NBER

NBER Law and Economics Summer Institute, Cambridge, July 28 2011.

Research question

Justin Wolfers, Comments on Matsa & Miller 2

� What are the effects of Norway mandating women on boards during a global financial crisis?

� What are the effects of mandating women on boards?

� Is there a female style in corporate leadership?

Evidence and Persuasion

Justin Wolfers, Comments on Matsa & Miller 3

Ho

w p

ers

ua

siv

e t

he

p

ap

er

is

How big the coefficient is

Bayesian

Popperian

Skeptical economist

Causal estimates of the effects of the mandate:

Justin Wolfers, Comments on Matsa & Miller 4

On firm outcomes:

� Return on Assets (=operating profit / assets): -4.0%� Mean: +3.4%

� Standard deviation: 17.5%

� Employment: +35% � All through less firing

On board composition

� Female share of board directors: +17%Interpretation:

� Estimating causal effects of one more female director

� Instrumental variables interpretation

� Wald estimator = ������������� �����

����������������������

� Implied effects of all-female boards is huge

How big is the experiment?

Justin Wolfers, Comments on Matsa & Miller 5

Mandate requires appointing women to boards

But not to the executive committee

Spain passed a similar law in 2007

France just passed a similar law

What’s missing?

Justin Wolfers, Comments on Matsa & Miller 6

� An analogy: What if they studied Saudi Arabia instead of Norway?

Oil

Justin Wolfers, Comments on Matsa & Miller 7

Annual Oil Production

Justin Wolfers, Comments on Matsa & Miller 8

0 100 200 300 400 500

QatarKuwait

Equatorial GuineaUnited Arab Emirates

NorwayBrunei

Saudi ArabiaLibyaOmanGabon

Virgin IslandsTrinidad & Tobago

AngolaAzerbaijan

CanadaVenezuela

KazakhstanIraq

East TimorRussia

BahrainCongroAlgeria

IranDenmarkEcuador

TurkmenistanMexico

MalaysiaAustralia

USAUK

Annual barrels per capita

Footnote 19

Justin Wolfers, Comments on Matsa & Miller 9

“In a related exercise we also

confirmed that the results are

unchanged if we exclude

companies in the petroleum

industry from the analysis.”

The logic of diffs-in-diffs-in-diffs

Justin Wolfers, Comments on Matsa & Miller 10

� Differencing strategy works if confounding unobserved factors are:

A. Common to the listed and unlisted sectors in Norway, Sweden and Finland

B. And have the same effect on the relative performance of the listed and listed sectors in Norway, Sweden and Finland

� Questions:

� Are the shocks common?

� Are the effects similar?

Labor economists assume yes (differencing)

Finance economists don’t—estimating firm-specific �’s

� Are the confounding shocks small?

Oil Prices

Justin Wolfers, Comments on Matsa & Miller 11

$0

$20

$40

$60

$80

$100

$120

$140

$160

19

87

19

89

19

91

19

93

19

95

19

97

19

99

20

01

20

03

20

05

20

07

20

09

20

11

Oil

Pri

ce (

$ p

er

ba

rre

l) Before After

Stock Prices

Justin Wolfers, Comments on Matsa & Miller 12

0

50

100

150

200

250

300

Sto

ck P

rice

In

dic

es

(20

03

=1

00

)

Before After

Norway

Sweden

Finland

General observation

Justin Wolfers, Comments on Matsa & Miller 13

In the wake of the global

financial crisis, we need to be

particularly careful about

intertemporal comparisons.

Ceterus ain’t parribus