The Political Foundations of Scarce and Unstable Credit

Charles W. Calomiris and Stephen Haber

Atlanta Fed ConferenceApril 9, 2013

Fact 1: Systemic Banking Crises are Endemic

Figure 1 The Frequency of Systemic Banking Crises, 1970-2010

Two or More Crises

No Crises

One Crisis 50%

Fact 2:Scarce credit is not randomly distributed (note the

relationship between stable democracies and credit provision)

Private Bank Credit as Percent of GDP, Average 1990-2009, by Income Groups

Libya

Colombia

U.K.

Upper MiddleIncome

Countries

Canada

USA

High Income Countries

Lower Middle Income Countries

Mean=29%

Low Income Countries Mean=12%

Mexico

0%

20%

40%

60%

80%

100%

120%

140%

160%

180%

Cong

o D

RCh

adUga

nda

Mal

awi

Nig

erRw

anda

Tanz

ania

Mad

agas

car

Moz

amH

aiti

Gam

bia

Buru

ndi

Togo

Mau

ritan

iaN

epal

Ango

laCo

ngo

Cam

eroo

nLe

soth

oSw

azila

ndIv

ory

C.G

uate

mal

aE.

Tim

orSr

i Lan

kaSa

moa

Phili

pIn

done

sia

Salv

ador

Boliv

iaM

oroc

coTo

nga

Jord

anLi

bya

Alge

riaBo

tsw

ana

Peru

Surin

ame

Dom

. Re

p.Co

sta

Rica

Colo

mbi

aBr

azil

Mau

ritiu

sS.

Afr

ica

Mal

aysi

aSa

udi A

.Tr

inid

adBa

hrai

nG

reec

eBa

rbad

osBr

unei

Mac

aoSw

eden

Isra

elBe

lgiu

mIc

elan

dSi

ngap

ore

Mal

taD

enm

ark

Spai

nN

. Ze

alLu

xem

U.K

.Cy

prus

Switz

Source: Calculated from World Bank Financial Structure Database (2010).

How many crisis-free, abundant-credit countries? What attributes shared?

• Singapore• Malta• Hong Kong, China• Australia• Canada• New Zealand

Half of these are small island or city states. The other half are democracies that have a history of

anti-populist constitutions.

How many high crisis, low credit countries are there--and what do

they have in common?High crisis, especially low credit: Chad, Democratic Republic

of the Congo.

High crisis, low credit: Argentina, Bolivia Brazil, Cameroon, the Central African Republic, Colombia, Costa Rica, Ecuador, Kenya, Mexico, Nigeria, the Philippines, Turkey, and Uruguay.

How many of these 16 countries have been stable democracies since 1970? Only 2: Colombia, Costa Rica.

Being a democracy is not the solution to endemic banking crises

Number of Systemic Banking Crises Since 1840, Canada and the USA

12

00

2

4

6

8

10

12

14

USA Canada

Among the puzzles we want to understand…

1. Why does there seem to be a relationship between autocratic government and weak banking systems?

2. Why are some democracies more prone to banking crises than others?

3. Why do government financing needs affect banking structure so differently in different countries?

The key to the answer: In order for there to be a banking system, three property rights problems have to be mitigated

1. Majority shareholders, minority shareholders, and depositors must be protected from expropriation by the government.

2. Depositors and minority shareholders must be protected from expropriation by majority shareholders.

3. Majority shareholders, depositors and minority shareholders must be protected from expropriation by debtors.

Solving these problems requires government, but governments have

inherent conflicts.

1. They simultaneously borrow from banks and regulate them.

2. They enforce debt contracts but need the political support of debtors.

3. They distribute losses in the event of bank failure, but they need the political support of depositors

Implications1. Banking systems are implicit partnerships between

governments and private actors.

2. That partnership is the product of a strategic interaction we call the “Game of Bank Bargains.”

3. The Game of Bank Bargains operates according to the logic of politics, not the logic of efficiency.

4. The game governs entry and competition, the pricing of credit and its terms, and the allocation of losses when banks fail.

5. Who is in the partnership varies across countries and within countries over time--because who is in the partnership depends on who is politically crucial.

A basic taxonomy of regimes and banking systems

Regime Government Banker-Government Partnership Banking System Outcomes

Chaos None None None No State

Absolute Power None None Poverty Trap

Centralized Rent Creating & Rent Narrow Credit, Strong StateSharing Network Locally Stable

AutocracyWeakly Centralized Inflation Tax Sharing between "Float" Banking Mid-Strength State

Oligarchy and Autocrat Local Oligarchies Little or No National Chartering Small, Fragmented Weak State

Liberalism Competitive Banking with Taxation Broad Credit, Stable Powerful State

DemocracyWelfare State Moves Banks Limited role for Banks Mid-Strength State

Populism out of the Line of Fire

Politically Determined Credit Broad Credit, Unstable Powerful State

Figure 1.1A Taxonomy of Regimes and Banking Systems

To show how political institutions and banking systems co-evolve we…

Look at what actually happened in five countries from the late 17th century to the present

The United Kingdom

The United States

Canada

Mexico

Brazil

Our five cases allow us to show each of the possible states of the world

England: Initially a crony system based on rent-sharing; later a system based on competitive banking with taxation.

The USA: Initially based on crony rent sharing; but dominated by populist banking since the 1820s.

Canada: Competitive banking with taxation.Mexico: No banking at all until 1880s, then crony banking

until 1990s; increasingly competitive and stable since democratization.

Brazil: Inflation tax banking from 1808 to 1994; increasingly competitive and stable since democratization.

England vs. Scotland

• Glorious Revolution and William III’s battle against Louis XIV…continuing war with France until 1815.

• Financing needs of sovereign defined and constrained English banking until 1825.

• But not Scotland (most innovative bank system in the world – branching, note clearing, lines of credit, interest-bearing deposits – as well as abundant credit deployed in diverse uses, lower risk of failure).

• Expansion of suffrage in 1832, Bank reforms of 1826, 1833, 1844 make England’s banks like Scotland’s.

The United States

A brief history of the U.S. Banking System.1. Populism versus liberalism at the outset. 2. The government--large banker partnership

of the early republic.3. Hamilton’s undoing. The unit banker-

agrarian populist partnership of 1830-1980.4. The government--large banker--urban

populist partnership of 1980 to the present

Populist outcome in the U.S.: 27,000 banks, but almost no branches

The Peculiar Structure of the U.S. Banking System, 1920 (A System of Tens of Thousands of Banks, but almost no Branches)

0

5,000

10,000

15,000

20,000

25,000

30,000

Banks Branches

Nu

mb

er o

f B

anks

an

d B

ank

Bra

nch

es

The death of the unit banker-agrarian populist coalition

The Number of Banks and Branches in the USA, 1920-2010

Number of Banks

Ratio of Branches per Bank (right axis)

0

5,000

10,000

15,000

20,000

25,000

30,000

1920

1923

1926

1929

1932

1935

1938

1941

1944

1947

1950

1953

1956

1959

1962

1965

1968

1971

1974

1977

1980

1983

1986

1989

1992

1995

1998

2001

2004

2007

2010

Num

ber o

f Ban

ks

0

2

4

6

8

10

12

14

Rat

io o

f Bra

nche

s Pe

r Ban

k

What changed?

• The agrarian-unit bank coalition was quite robust – it survived the Civil War, the banking reform movement c. 1910 in response to the Panic of 1907, and the Great Depression.

• Five influences that unwound it: (1) demography, (2) technology (ATMs) and court decisions, (3) domestic disintermediation, (4) loss of global market share, and (5) crises of 1980s.

How did we get to a state of the world in which anyone could get a mortgage?

Figure 8.2 High Risk Mortgage Originations, U.S. Market, 2001-2006 (in $ Billions, High Risk=Subprime plus Alt-A Mortgages)

$1 Trillion

Subprime plusAlt-A

Originations

34%Subprime plus Alt-A

as % of Total (right axis)

$0

$200

$400

$600

$800

$1,000

$1,200

2001 2002 2003 2004 2005 2006Source: Acharya et al., (2011), p. 46.

0%

5%

10%

15%

20%

25%

30%

35%

Subp

rime

plus

Alt-

A M

ortg

age

Orig

inat

ions

, as

Perc

ent o

f All

Mor

tgag

e Le

ndin

g

In which it was possible to borrow without collateral?

Figure 7.4 Percent of Home Purchases in the United States with a

Downpayment of Three Percent or Less, 1980-2007

0.3% 0.5%

10.0%

14.3%

40.0%

0%

5%

10%

15%

20%

25%

30%

35%

40%

45%

1980 1990 1999 2003 2007Source: Pinto (2011), pp. 24-25.

Part of the answer is financing by the GSE’s--but that just begs the question

Figure 8.1 The Growth of GSE High Risk Mortgages, 2003-2007

(New GSE Mortgage Purchases, High Risk=LTV>80 and/or FICO<660.)

36%

GSE New High Risk, as % Total NewGSE Lending

87%

GSE New High Risk, as % Total New High Risk Lending

0%

5%

10%

15%

20%

25%

30%

35%

40%

2003 2004 2005 2006 2007

Source: Acharya et al., (2011), p. 59.

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

GSE

Hig

h R

isk,

as

% T

otal

Hig

h R

isk

The answer is rooted in the political institutions governing the creation of too big

to fail megabanks“We support the NationsBank acquisition of BankAmerica

because…they will make credit work for low and moderate income people and they will work with the community institutions.”

“NationsBank… [was] the first multistate lender to negotiate their mortgage underwriting standards with us…. At the time these things were pretty radical, but today no one thinks twice about the appropriate use of low downpayments, nontraditional credit, food stamps as income, voluntary child support, cash on hand, or steady income rather than the same job for two years.”

--George Butts, President of ACORN Housing, from his testimony to the Federal Reserve Board in support of the acquisition of BankAmerica by NationsBank, July 9 1998.

The curious coalition formed by merging banks and activist groups

Figure 7.2 Cumulative Value of CRA Agreements Between

Banks

$14 Billion

$579 Billion

$867 Billion

$0

$100

$200

$300

$400

$500

$600

$700

$800

$900

$1,000

1977

1978

1979

1980

1981

1982

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

Source: National Community Reinvestment Coalition, CRA Commitments (2007), pp. 11-17.

These deals only capture a subset of all CRA transactions

Figure 7.1 Cumulative CRA Commitments, 1977-

2007

$43 Billion

$1.3 Trillion

$4.6 Trillion

$0

$500

$1,000

$1,500

$2,000

$2,500

$3,000

$3,500

$4,000

$4,500

$5,000

1977

1978

1979

1980

1981

1982

1983

1984

1985

1986

1987

1988

1989

1990

1991

1992

1993

1994

1995

1996

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

Source: National Community Reinvestment Coalition, CRA Commitments (2007), p. 8.

That coalition became further institutionalized through the GSE’s

Figure 7.3 HUD Loan Repurchase Mandates for Fannie and Freddie, 1992-

Low and Moderate Income Lending

Special Affordable (very low income)Lending

Underserved Area Lending

0%

10%

20%

30%

40%

50%

60%

1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008

Source Pinto (2011), pp. 76, 87, 104, 105. Note: Fannie and Freddie actual loan purchases met these goals.

The outcome is well-known

Figure 8.2 US and Canadian Mortgages in Arrears, 1991-2011

Canada

USA

0%

1%

2%

3%

4%

5%

6%

7%

8%

9%

10%

11%

1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011Source: U.S. Federal Reserve; Canadian Bankers Association. In the U.S., mortgages are in arrears after 30

days, in Canada after 90 days. Thus, the most meaningful comparison is the ratio of the two over time.

Why are Canadian political institutions so different from those of the U.S.?

Implications for Canada’s government-banker partnership

1. Banking policy in Canada was centralized in the national government; while banking policy in the US was historically left to the states. Also, Senate is appointed, and voting is non-proportional.

2. It was therefore possible to build local rent seeking coalitions in the US, but not in Canada.

3. In Canada, populist banking proposals are always beaten back: Canada has had a system of a few large banks since 1817.

4. The limits imposed on the Canadian partnership by the franchise (the unusual nature of Canadian bank charters).

Three Main Conclusions1. Banks can only develop with the active encouragement of

government, and they can only provide credit if government financing needs permit them to do so (England vs. Scotland).

2. In populist democracies, such as the U.S., bank regulation is used as a political tool to favor some parties over others. Instability is tolerated by controlling coalition as the price for obtaining the benefits that it extracts from controlling banking regulation. In countries where it is difficult to form coalitions to give special access to bank credit to a subset of the population, stable and abundant bank credit – are much more likely.

3. Regulatory failures – e.g., the decision not to permit nationwide branch banking in the United States prior to the 1990s, or the Fed’s permissive approach to bank mergers in the 1990s –are outcomes of political bargains. This implies that the prospects for regulatory reform are often bleak.