No.

62
J uly 2009
This study was
prepared under the
authority of the
Treasury and
Economic Policy
General Directorate
and does not
necessarily reflect the
position of the M inistry
for the Economy,
I ndustry and
Employment.
How does today's US crisis compare with
the 1990s J apanese crisis?
Both the Amer ican and the Japanese cr ises or iginated in the bur sting of specula-
tive bubbles, for cing pr ivate agents households in the case of the USA, and non-
banks in the Japanese case to r educe their debt. In the case of Japan, debt
r eduction in the midst of a financial cr isis tr igger ed a deflationar y spir al. A Japa-
nese-style deflationar y spir al seems unlikely in the United States, despite simila-
r ities r egar ding the depth of the financial cr isis and the scale of the excesses
needing to be unwound.
Japan enter ed a deflationar y spir al in the wake of a pr otr acted depr ession as a
r esult of a ser ies of specific factor s, some inter nal ( such as the length of time
taken to br ing hidden doubtful assets out into the open) , and some exter nal ( e.g.
the Asian cr isis and the appr eciation of the yen) , at a time when the economy was
alr eady extr emely fr agile.
A highly aggr essive economic policy r esponse could enable the United States to
aver t a deflationar y spir al. The Amer ican author ities ar e concentr ating on avoi-
ding r epeating the mistakes of their Japanese counter par ts. Their management
of the cr isis appear s to have been mor e r esponsive ex ante, and mor e ambitious
in scope, with a ser ies of r apid, tar geted and lar ge-scale stimulus plans, aggr es-
sive r ate-cutting, unconventional monetar y policy measur es, and the cr eation of
a defeasance str uctur e to buy-up "toxic assets".
The dr op in consumer pr ices on a
year -on-year basis obser ved in the
United States since Mar ch 2009 is
mainly attr ibutable to a base effect on
ener gy pr ices. This is likely to be tem-
por ar y only.
Source: DGTPE
Monetary policy responses after the bursting of the US and J apanese
property bubbles
0
1
2
3
4
5
6
7
8
9
- 2 yrs property
bubble bursts
2 yrs 4 yrs 6 yrs 8 yrs 10 yrs
Japanese key rate Fed key rate
%
Fed implements
unconventional
monetary policy
BoJ implements
quantitative easing
TRSOR-ECONOMICS No. 62 J uly 2009 p. 2
1. The origins of the American crisis resemble those of the J apanese crisis in the 1990s
1.1 J apan's deflationary crisis arose fromnon-
banks' need to reduce their debt
Pr ior to the Japanese cr isis ther e was an abundance of liqui-
dity caused by the r apid easing of monetar y policy as inflatio-
nar y pr essur es abated in the 19 80s, and by the shar p r ise in
the supply of bank lending br ought on by incr eased compe-
tition among the banks. This led to the for mation of stock
mar ket assets and pr oper ty bubbles, which bur st following a
r ise in Japanese and global inter est r ates. In par ticular , non-
banks had taken advantage fr om cheap capital and the steep
r un-up in their mar ket valuations to r aise their bank
bor r owings, which significantly suppor ted gr owth via an
appr eciable r ise in investment. These bor r owings wer e faci-
litated by the special natur e of Japan' s pr oductive set-up,
featur ing extr emely str ong links between the industr ial
car tels ( the Keir etsu) and the financial system.
The pr oper ty and stock mar kets fell shar ply after the bubble
bur st, r ender ing the banking system insolvent. The banks,
which have histor ically been r elatively unpr ofitable, wer e
slow to clean up their balance sheets in or der to limit their
losses. Bank liquidity and the supply of cr edit dwindled as a
r esult. A non-optimum allocation of lending to unpr ofitable
fir ms
1
thus set in, while sound fir ms found themselves
star ved of liquidity. In addition, the decline in the pr ice of
pr oper ty and stock mar ket assets led to a slump in company
pr ofits and, combined with shr inking cr edit, pushed mor e
and mor e fir ms into bankr uptcy. The deter ior ation in
company and bank balance sheets, together with a lar ge
number of bankr uptcies and the tight links between compa-
nies and banks pr ompted a steep r ise in doubtful loans,
which squeezed banks' balance sheets and fur ther r estr icted
the supply of cr edit ( see Char ts 1 and 2) .
Non-banks needed to r educe their debt r apidly, their finan-
cial condition having deter ior ated sever ely in the ear ly-
1990s
2
. As a r esult, companies shar ply cut their investment
spending ( see Char t 3) , ther eby r educing their demand,
weakening the Japanese economy, and putting downwar d
pr essur e on pr ices. Consequently the r oots of Japan' s defla-
tionar y cr isis lie in companies' need to pay down their debt,
not in any weakness in household consumption. Indeed,
households have dr awn down their savings since the ear ly-
1990s
3
, ther eby compensating for the fall in cor por ate
sector demand, sustaining domestic demand and counter ing
deflationar y pr essur es.
Chart 1: J apan: doubtful loans and bank losses from doubtful loans
Source: FSA
Chart 2: J apan: bank lending to private agents
Source: OECD
Chart 3: J apan: debt and investment by private-sector non-banks
Source: Bank of Japan, National Accounts
1.2 In the United States, the same debt deflation
mechanismposes a risk, but in this case it con-
cerns households
As in Japan in the late-1980s, a similar expansion of liquidity
was obser ved in the United States fr om ar ound 2000
onwar ds. Her e, though, it was caused by highly accommoda-
(1) The banks tended to come to the rescue of defaulting borrowers to enable the latter to pay the interest due on their
debt and thus avoid the need for the banks to classify these loans as in default. As a result, bank lending was allocated
primarily to the least profitable firms, thus reducing the supply of credit to healthy companies and helping to keep
unprofitable firms in business. See "La dflation japonaise: le rle-clef du besoin d'ajustement des bilans des
entreprises" (Japanese deflation: the key role played by firms' need to adjust their balance sheets). Diagnostics Prvisions
Analyses conomiques no. 29 -February 2004.
(2) The massive investment of the late-1980s led to a huge increase in borrowing requirements, characteristic of an
economy emerging from an investment bubble.
(3) There are several possible explanations for this downtrend in the saving rate, including population ageing, a
"substitution" effect favoured by lower interest rates, and the substitution of consumption for investment in housing.
0.0%
0.5%
1.0%
1.5%
2.0%
2.5%
3.0%
0%
1%
2%
3%
4%
5%
6%
7%
8%
9%
1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
Doubtful loans held by all high street banks (lefthand scale)
Bank losses due to doubtful loans
% OF GDP %of GDP
most recent data point: 2005
-8%
-6%
-4%
-2%
0%
2%
4%
6%
8%
10%
12%
14%
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006
Year-on-year
most recent data point: December 2006
70%
80%
90%
100%
110%
120%
130%
140%
12%
13%
14%
15%
16%
17%
18%
19%
20%
21%
1980 1982 1984 1986 1988 1990 1992 1994 1996 1998 2000 2002 2004 2006
ratioof investment to GDP gross debt of companies (righthand scale)
Bursting of stockmarket bubble
most recent data point: 2007
%of GDP
TRSOR-ECONOMICS No. 62 J uly 2009 p. 3
tive monetar y policy, accentuated by the steep r ise in global
liquidity since the beginning of the 2000s.
The for eign exchange r eser ves accumulated by the Asian
countr ies as their economies enjoyed r apid gr owth wer e
indeed absor bed by the United States. As in Japan, abundant
liquidity dr ove up the pr ices of financial and pr oper ty assets,
thus bolster ing bor r ower s' appar ent solvency and encour a-
ging fur ther lever age. This in tur n fuelled demand for assets
and, ultimately, pushed up their pr ices still fur ther , leading
to high levels of indebtedness and investment.
The r esidential pr oper ty bubble in the United States appear s
to have been compar able in scale to that of Japan. While r esi-
dential sector pr oper ty asset pr ices r ose less br utally in the
United States, they have fallen slightly faster than in Japan; the
decline seems to be similar in the commer cial sector , for the
moment, even if the United States is cur r ently r egister ing only
its fir st quar ter s of falling pr ices. On the other hand, Japan' s
financial asset pr ice bubble was mor e than ten times gr eater
than in the United States. Japanese stock mar ket assets lost
35% of their value in the space of a year following the bur s-
ting of the bubble. In the United States, meanwhile, 18
months after peaking asset pr ices lost near ly 40% of their
value
4
. This decline r eflects the weakness of the financial
system, the collapse of major mar ket player s having pr ecipi-
tated a steep fall in asset pr ices and incr eased volatility.
Chart 4: Asset prices before and after the bursting of the bubble in J apan
and the fall in asset prices in the United States
Source: Datastream, DGTPE calculations
Chart 5: Residential property asset prices before and after the bursting of
the bubble in the United States and J apan (national price indices)
Sources: FHFA (United States), Japan Real Estate Institute (Japan),
DGTPE calculations
The speculative excesses needing to be cor r ected differ ed
between the two countr ies. In the United States, easier access
to cr edit ( fuelled by expectations of r ising pr oper ty pr ices
and the emer gence of financial pr oducts designed to r educe
r isk exposur e) led to a shar p expansion of the mor tgage
mar ket. This tr igger ed a massive r ise in household debt,
without affecting investment by non-banks, which appear to
be in sound financial condition.
The US pr oper ty pr ice bubble bur st as a r esult of slowing
demand for mor tgage loans. This, together with r ising
defaults
5
, r ever sed pr ice expectations, leading finally to a fall
in pr ices. Just as pr oper ty pr ices began falling, this loan
financing mechanism gr ound to a halt. The pr oper ty cr isis
then fed thr ough to the financial sector via r ising defaults on
subpr ime loans in the mor tgage mar ket. The r ise in house-
Box 1: How deflation works
A deflationary spiral, where falling prices and demand become a self-sustaining process, is possible only if at least one of the following
three factors is at work:
Falling inflationary expectations
Households tend to postpone consumption if they think prices are going to continue to fall. This kind of wait-and-see attitude ultimately
weakens demand for goods and services provided by companies. This will lead firms to cut their output, and hence their demand for labour
and/or the wages they pay their employees. Similarly, if companies expect prices to fall, they will expect their profits to fall too; to limit the
erosion of their margins, they will reduce their demand for labour and/or cut wages. Unemployment rises and wages fall, amplifying the
economic slowdown and the fall in prices. Ultimately, falling prices and domestic demand feed on each other.
Certain over-indebted agents see their real debt rise as asset prices fall: this is known as debt-deflation
The real cost of debt rises when the price of real or financial assets falls steeply. Faced with the risk of insolvency, agents may be forced to
reduce their debt leverage. But falling collateral values leads to a drying up of the supply of credit. Sources of financing and loan repay-
ment dwindle. Borrowers seek to sell off their assets to avoid bankruptcy, further undermining asset prices. But falling prices in turn lead to
rising defaults and amplifies the rationing of bank credit. Consequently, domestic demand declines sharply leading to a fall in the general
price level.
The economy is in a liquidity trap
Here, nominal interest rates have reached a floor and can fall no further. Falling prices therefore lead to a rise in real interest rates and
monetary policy ceases to be effective as a tool for stimulating the economy. Rising interest rates tend to weaken final demand by raising
the cost of borrowing and aggravating the burden on borrowers.
In J apan, over-indebtedness was the main mechanism triggering the deflationary process, which was subsequently sustained by expecta-
tions and the liquidity trap.
(4) Japan's stockmarket continued its downward course in the 1990s, and had lost 60% of its value eight years after the
bursting of the asset bubble.
45
55
65
75
85
95
105
- 3yrs - 2 yrs - 1 yr bubblebursts 1yr 2yrs
United States (S&P500) Japan (Nikkei)
Base 100 at peak
(5) The rise in default rates in 2006 was partly the result of dearer than expected subprime loan repayments (the interest
rate on repayments is low at the start of the loan, then rises subsequently); this was compounded by the rise in the key
rate, which fed through into mortgage rates, thus strangling weaker borrowers who had borrowed at variable rates.
80
85
90
95
100
105
- 2 yrs - 1 yr and 6
mths
- 1 yr - 6 mths bubble
bursts
6mois 1yr 1yr and 6
mths
2yrs
United States (property prices) Japan (land prices)
Base 100 at peak
TRSOR-ECONOMICS No. 62 J uly 2009 p. 4
hold defaults entailed hefty losses for financial institutions,
spr eading to the entir e financial system via str uctur ed
pr oducts and "r epackaged" loans; in some ways this was
much the same mechanism as the one at wor k with doubtful
loans in Japan. Lack of tr anspar ency as to financial institu-
tions' r eal exposur e and losses br ed a cr isis of confidence in
the inter bank mar ket. This fir st mater ialised in a dr ying up
of liquidity in this mar ket, followed by a spate of failur es
among financial institutions ( most pr ominently Lehman
Br other s) . These failur es sever ely destabilised the mar kets,
cr eating additional difficulties for the banks. A vicious cir cle
then ar ose in the financial mar kets leading institutions to
delever age extensively and r ein-in their lending. The lack of
tr anspar ency as to institutions' r eal balance sheet exposur e
and losses pr ompted a wave of distr ust that sever ely
impair ed the wor kings of the global financial system, as US
mor tgage-r elated r isk spr ead wor ldwide.
Thus, the r oots of the Amer ican and Japanese cr ises wer e
basically the same, so we can imagine them having the same
deflationar y consequences. However , the deflationar y
mechanisms at wor k in the two countr ies ar e pr obably ver y
differ ent, due in par ticular to the specific modes of adjust-
ment of the labour mar ket ( in the United States, depr essed
final demand is channelled via the adjustment of employ-
ment and not via a pr ice-wage loop) and the inter national
envir onment ( the yen' s appr eciation in par ticular exer ted
additional downwar d pr essur e on pr ices) .
Chart 6: US household debt, as a % of gross disposable income (GDI)
Source: Fed
Chart 7: Private non-financial agents' debt, as a % of GDP
Sources: BoJ, Fed, DGTPE calculations
2. The main differences in the deflationary mechanisms in J apan and the United States lie in the way their labour
markets and the international environment adjust
The deter ior ation in the inter national envir onment, which
accentuated the Japanese slowdown, had a cause exogenous
to the Japanese cr isis, namely the Asian cr isis of 1997, which
sever ely hur t Japanese expor ts ( the Japanese economy' s
main dr iver ) . The cur r ent collapse in global demand for US
goods and ser vices, meanwhile, stems mainly fr om the
contagion of the Amer ican cr isis to the r est of the wor ld. The
United States, which was at the or igin of the cr isis, has r egis-
ter ed a ver y shar p contr action in impor ts due to the collapse
of domestic demand. This decline in impor ts has mor e than
offset that of expor ts caused by weaker global demand. The
US tr ade deficit has shr unk dr astically. Over all, for eign tr ade
made a positive contr ibution to US GDP gr owth in 2008 and
in the fir st half of 2009. Between 1997 and 1999, on the
other hand, because the slowdown in global demand for
Japanese goods and ser vices was exogenous to the Japanese
economy, expor ts weakened so sever ely that for eign tr ade' s
contr ibution to gr owth was nil.
The yen' s successive appr eciations in the cour se of the
1990s had alr eady depr essed pr ices still fur ther ( via
impor ts) and hamper ed expor ts as a r esult of r educed
competitiveness. This phenomenon was exacer bated by the
fall in global demand for Japanese goods and ser vices dur ing
the Asian cr isis. The United States, on the other hand, is not
suffer ing fr om an adver se cur r ency movement, since the
dollar weakened when the US enter ed r ecession in the four th
quar ter of 2007, helping to sustain activity.
In Japan, the labour mar ket adjustment in r esponse to
shar ply r educed domestic demand and, ultimately, GDP,
occur r ed via wages, wher eas in the United States the bur den
appear s to be falling mor e on jobs. In Japan, wages ar e
downwar dly flexible in times of cr isis: Japanese tr ade unions
have an implicit total employment tar get and hence will
agr ee to wage cuts when th e economy slows, to pr event
unemployment fr om r ising unduly ( unemployment did not
exceed 3. 5% until the end of the 1990s) . Consequently,
lower bonuses and r educed over time wor king tr aditionally
ver y substantial was one of the main lever s activated by
employer s to adjust wages downwar ds. Wages had alr eady
begun to slow at the beginning of the cr isis, but the downtur n
in activity in 1997 and its impact on domestic demand this
time led them to fall significantly.
This exer ted negative pr essur e on pr ices ( see Char t 8) ; these
had alr eady come under deflationar y pr essur e with the
opening up of emer ging Asian countr ies
6
and their falling
expor t pr ices.
0%
20%
40%
60%
80%
100%
120%
140%
1990 1992 1994 1996 1998 2000 2002 2004 2006 2008
Aggregate Mortgage Consumer borrowing
Most recent data point: Q1 2009
0%
50%
100%
150%
200%
250%
Japan United States
Q-10ans Q(bursting of strockmarket bubble)
(6) Deregulation opened up markets to foreign manufactures, whose share of total imports rose from 23% to 59%
between 1980 and 1995.
TRSOR-ECONOMICS No. 62 J uly 2009 p. 5
Chart 8: Wages and inflation in J apan
Sources: Datastream, DGTPE calculations
In the United States, on the other hand, wages ar e bar ely
cyclical at all, and have not r eally slowed since the onset of
the cr isis. Employment has been the labour mar ket' s adjust-
ment var iable, her e. Consequently, while a pr ice-wage loop
looks less likely in the United States than in Japan, r ising
unemployment could never theless depr ess final demand and
weigh on pr ices indir ectly. The main differ ence is that the
deflationar y mechanism could be less dir ect in the United
States than in Japan.
Ther e is, mor eover , a fundamental deflationar y component
in Japan that is appar ently absent in the United States, namely
an economic policy inappr opr iate to the sever ity of the cr isis.
Chart 9: Inflation and the labour market in the United States
Source: BLS
3. US policy appears better suited than J apanese policy to countering the risk of deflation
3.1 The J apanese authorities were slow to react
The Bank of Japan ( BoJ) went on r aising its r ates after the
financial bubble bur st, which helped to weaken the banking
system. Not until the explosion of the pr oper ty cr isis ( one
year after that of the stockmar ket bubble) did it begin to
lower its key r ates, and it took mor e than five year s to cut
them to 0.5%. Rate cutting pr oved insufficient to avoid a
wave of bankr uptcies among financial institutions, alr eady
weighed down by heavy losses.
However , this monetar y policy pr oved ineffectual, since the
cr edit channel was blocked ( the banks' balance sheets wer e
too weak for a cr edit-led stimulus. Cr edit r ationing set in
fr om 1998 onwar ds, while negative inflation dr ove up r eal
inter est r ates, cr eating a liquidity tr ap that fur ther bolster ed
the deflationar y spir al ( see Box 1) . This situation stemmed
lar gely fr om the length of time the banks, the gover nment
and the BoJ
7
took to acknowledge the impor tance of the
question of doubtful loans and addr ess it.
-6%
-4%
-2%
0%
2%
4%
6%
8%
1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
wages (y-o-y quarterly moving average inflation (y-o-y)
VAT is increased
0
1
2
3
4
5
6
7
8
9 -1
0
1
2
3
4
5
6
7
1990 1992 1994 1996 1998 2000 2002 2004 2006 2008
inflation per capita average wage unemployment rate (righthand scale)
Y-o-y in % %
Most recent data point: Q1 2009
Box 2: J apanese crisis timeline
Source : DGTPE

+2 yrs
1
s t
fi scal
st imulus
Heavy banking
losses and ba nk
fail ur es + ri si ng
unemployment
Act ivi ty slows Pr ices sl ow
+3 yrs
+1 yr
Wave of
cor porat e
bankr uptcies
Propert y bubbl e burst s
Key r ates cutti ng
Recession
+8 yr s
(1998)
Sev ere fi nancial cri sis :
banks nati onal ised ; fi rst
measures to rescue
fi nancial system
Dual shock: tax
shock and Asi an
cri si s
Financi al bubbl e
bursts
Q4
1989
Deflat ion
BoJ pursues zer o
inter est r ate pol icy
Phase 1: Economy weakens , 19901997
Phase 2: Deflation
takes hold, 19981999
Ph as e 3: Defl ati onary spi ral
20002003
+10 yr s
BoJ i mplements
unconventi onal
monetary pol icy
(7) Japanese law lays down three definitions for doubtful loans. The first, and the first associated data, did not come into
being until 1993 and after, while the other two definitions were not introduced until 1999, following the "Financial
Reconstruction Bill", i.e. nine years after the onset of the crisis (see "Non performing loans and the real economy:
Japan experience", Inaba et al, 2005 - Bank for International Settlements 22-07 pages 106-27).
TRSOR-ECONOMICS No. 62 J uly 2009 p. 6
The pr oblem did not come to the for e until Japan was on the
br ink of systemic cr isis. This was symptomatic of the Japa-
nese cr isis and is one of the key factor s contr ibuting to its
extr aor dinar y dur ation.
The Japanese gover nment unveiled ten stimulus plans
between 1992 and 2000, for a cumulative amount r epr esen-
ting ar ound 27% of GDP and split r oughly evenly between
r eal budgetar y spending and financial measur es such as life-
lines for the banks and cr edit facilities for SMEs. For a var iety
of r easons, however , these plans pr oved ineffectual in view
of their official size ( see Box 3) .
Chart 10: J apan's monetary policy response to the bursting of the asset
price bubbles
Source: OECD
So it looks as if the appr opr iate economic policy was not
applied until too late: ex post, the Fed' s models
8
show that a
mor e significant r ate cut befor e 1995 and a mor e accommo-
dating fiscal policy would have pr evented Japan fr om falling
into deflation. It has only been possible to establish the
belated natur e of the BoJ' s action ex post. Ex ante, in their
ignor ance of the banking system' s difficulties and compa-
nies' fr ailty, together with over -optimistic for ecasts of infla-
tion and activity, analysts consider ed the monetar y policy
r esponse to be appr opr iate.
It was not until 1998 and after war ds, r eally, that vigor ous
steps wer e taken to stem the financial cr isis ( see Table 1) .
Taken together with the BoJ' s zer o inter est-r ate policy and
quantitative measur es fr om 2001 onwar ds ( see Box 4) ,
these steps slowly succeeded in pur ging the mar ket of bad
debts and clear ing up the cr isis. However , the financial
system' s super visor y author ity failed to pur sue the bank
stabilisation measur es thr ough to their conclusion
9
, and it
was not until 2003 that the gover nment pr oceeded to buy up
all doubtful loans ( see Char t 1) . Zer o inter est r ate policies
followed by quantitative easing successfully modified agents'
expectations as to the time fr ame for futur e shor t-ter m r ates;
this did push down long r ates a little, but their ability to
r evive lending in Japan and shor e-up inflationar y expecta-
tions, on the other hand, was at best ver y slow to mater ialise.
Source: DGTPE
Indeed, the str ong gr owth in the monetar y base did not
pr oduce a sufficient incr ease in money supply, since the
banks did not alter their lending behaviour in consequence
( the monetar y multiplier being ver y weak) and expanded
their holdings of safe assets. The r elative ineffectiveness of
monetar y policy can also be accounted for by the fact that the
BoJ' s expansionar y policy lacked cr edibility. Indeed, in
2000, with deflation in full swing, the BoJ fir st r aised its
r ates, then cut them shor tly after war ds, and it was not until
2003 that the BoJ began communicating clear ly about how it
planned to exit fr om quantitative easing
10
.
0
1
2
3
4
5
6
7
8
9
1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002
Key rate
%
Financial bubble
bursts
Property bubble
bursts
Box 3: Why J apan's fiscal stimulus was ineffectual despite its scale
Of all of the J apanese government's stimulus plans (amounting to a cumulative 27% of GDP, roughly), growth-boosting measures repre-
sented only 1/3
a
of the total amount, on average (or a cumulative 10% of GDP, roughly). The macroeconomic effectiveness of these measu-
res was also impaired by lags in implementing them and their overly limited scopewhen not counteracted by restrictive counter-
measures. That is because one-off fiscal consolidation measures are reckoned to have counterbalanced the positive effects of stimulus
plans, plunging J apan into deflation in 1998. The tax shock of 1997 (VAT was raised from 3 to 5%, the income tax reductions in force since
1995 were reversed, the health service patient contribution was increased, and public investment was pruned sharply) contributed to the
J apanese recession in 1997/1998. In particular, over and above the temporary positive shock to inflation, the VAT rate hike pushed prices
into negative territory as consumption plummeted. Other possible explanations of these plans' ineffectiveness are that the measures were
poorly targeted (being aimed at underperforming sectors), and that a liquidity trap emerged.
a. "Fiscal policy works when it is tried", Adam Posen, Institute of International Economics, 1998.
(8) Alan Ahearne et al: "Preventing Deflation: Lessons from Japan's Experience in the 1990s," International Finance
Discussion Papers no. 729; 2002.
(9) What is more, the bank stabilisation measures implemented in 1998 and after were sub-optimal: two public liquidity
injection programmes were applied in parallel to the public deposit guarantee mechanism, but the counterparts
demanded by the authorities were never provided, which created moral hazard and did not encourage the banks to
speed the structuring of their assets, declare bankruptcy, or reveal their losses.
Table 1: J apan: economic and monetary policy measures
to rescue the financial systemfrom1998onwards, in
percentage points of 2000GDP
Public equity injections 2.5
Guar antees to banks 3.7
Cr eation of defeasance str uctur es to pur chase
doubtful loans ( in 1999 and 2003)
2.0
Gr anting of loans at differ ential r ates to tr oubled
banks
1.0
Pur chase by BoJ of shar es fr om banks 0.4
TOTAL 9.8
(10) The conditions for exiting from quantitative easing were that year-on-year inflation of non fresh-food prices (Core
CPI) had to be positive over several months and that forecasts should not be negative. Krugman explains that these
conditions were not sufficiently binding and that, for the expansionary monetary policy to be truly credible, the BoJ
would have had to adopt a 4% inflation target over 15 years, versus a 0% inflation target during the years of
quantitative easing: "It's back: Japan's slump and the return of the liquidity trap", Krugman, Brookings Papers on
Economic Activity, vol. 29 (1998-2) pages 137-206; 1998).
TRSOR-ECONOMICS No. 62 J uly 2009 p. 7
3.2 The United States has learned the lessons of
the J apanese crisis
3.2.1 A highly aggressive monetary policy
The Amer ican author ities appear to have managed the cr isis
mor e r esponsively and mor e ener getically. They began inter -
vening in the monetar y spher e str aight after the bur sting of
the pr oper ty and financial bubble, wher eas the BoJ had
waited until the bur sting of th e pr oper ty bubble, one year
after the financial bubble bur st. But r ight fr om the star t of
their monetar y inter vention, the United States and Japan cut
their r ates at a compar able pace ( r espectively 500 pb ver sus
460 pb over the following 21 months ( see Char t 7) .
However , the Fed took less than two year s to cut its r ate to
zer o, compar ed with eight year s in Japan, wher e the key r ate
had been higher when the bubble bur st..
As a stop-gap measur e to offset any br eakdown in financing
to the economy, the "financial system stabilisation plan" also
pr ovides for mor e dir ect measur es in the for m of loans to
banks and r ecapitalisations. But this plan also pr ovides
dir ect suppor t for lending to households and small busi-
nesses, contr ibuting mor e than $ 100 billion to the Fed' s
TALF pr ogr amme
11
. The Fed, meanwhile, has come up with
an ar r ay of new cr edit instr uments to help pur sue its mone-
tar y easing and keep the pr imar y and secondar y mar kets
liquid. This policy is descr ibed as cr edit easing ( see Box 4) .
Another of the Fed' s objectives was to br ing down long-ter m
r ates. The implementation of "unconventional" measur es
such as pur chases of Mor tgage Backed Secur ities, secur ities
of gover nment-sponsor ed enter pr ises ( GSE)
12
and US
Tr easur y bonds, br ought about a substantial fall in long-ter m
mor tgage r ates, to some extent holding in check the r ise in
10-year gover nment bonds, and a depr eciation of the dollar .
Although the situation appear s to have impr oved wher e liqui-
dity is concer ned, the imper ative need is to r estor e the cr edit
channel to nor mal once mor e, this being the main instr u-
ment of monetar y policy and the indispensable means of
financing to business and households.
Chart 11: Monetary policy response after the bursting of the J apanese and
US property bubbles
Sources: BoJ, Fed, DGTPE
The Fed' s intr oduction of a pur e money cr eation mechanism
ser ves to cur b deflationar y expectations, since these
measur es should have an inflationar y impact once activity
picks up. Agents' expectations could then shift fr om deflatio-
nar y to inflationar y, unless agents expect inter est r ates to r ise
once the cr isis is over , in which case the policy of money
cr eation would be ineffective as a means of anchor ing expec-
tations.
Box 4: Unconventional monetary policies in J apan and the United States
Between 2001 and 2006, more than a decade after the outbreak of the crisis, J apan took unconventional policies measures described as
quantitative easing, substituting a balance sheet size objective for an interest rate objective. This policy was liabilities-oriented, using open-
market operations to achieve a quantitative target (revised upwards several times) via the current accounts of the J apanese private-sector
banks on the liabilities side of the BoJ balance sheet. The central bank also expanded the modus operandi of its interventions to include
purchases of treasury bonds and, from 2002 onwards, equities, ABS and commercial paper. In addition, the Finance Ministry intervened
heavily in the foreign exchange markets in 2003 and 2004. While this policy did curb the yen's appreciation and provided cheap, abundant
funding to the banks, it does not appear to have any major impact on activity
a
.
The Fed's policy, described by its Chairman as credit easing, concentrates more on the central bank's assets. The Fed is seeking to act on
several segments of the capital markets (commercial paper, MBS, Treasury paper, etc.) by purchasing securities. This is swelling the central
bank's assets, the purchases being financed by the creation of money moreover (the central bank credits the accounts of the banks selling
these securities). The very large number of loans made against more-or-less high-quality collateral has sharply expanded the Fed's balance
sheet.
a. See Trsor Economics no. 56: Unconventional monetary policies, an appraisal, and IMF, 2009 Gauging risks for deflation.
(11) Term Asset Backed Securities Loan Facility: this programme supports the issuance of ABS (Asset Backed Securities)
collateralised by student and car loans, etc. and provides for loans to holders of certain triple-A rated ABS assets.
(12) Government Sponsored Enterprises are public/private entities specialising in mortgage loans to households generally
regarded as risky. The best-known GSEs are Fannie Mae and Freddie Mac.
0
1
2
3
4
5
6
7
8
9
- 2 yrs - 1 yr property
bubble bursts
1 yr 2 yrs 3 yrs 4 yrs
Japanese key rate USkey rate
%
Japanese
financial bubble
bursts
US financial
bubble bursts
TRSOR-ECONOMICS No. 62 J uly 2009 p. 8

Publisher:
M inistre de l conomie,
de l I ndustrie et de l Emploi
Direction Gnrale du Trsor
et de la Politique conomique
139, rue de Bercy
75575 Paris CEDEX 12
Publication manager:
Philippe Bouyoux
Editor in chief:
Jean-Paul DEPECKER
+ 33 (0)1 44 87 18 51
tresor-eco@ dgtpe.fr
English translation:
Centre de traduction des minis-
tres conomique et financier
Layout:
M aryse Dos Santos
I SSN 1777-8050
R
e
c
e
n
t

I
s
s
u
e
s

i
n

E
n
g
l
i
s
h

J uly 2009
No. 61. The Revenu de Solidarit Active or earned income supplement: its design and expected
outcomes
Clment Bourgeois, Chlo Tavan
J une 2009
No. 60. China, laboratory to the world?
Alain Berder, Franois Blanc, J ean-J acques Pierrat
No. 59. Distributable surplus and share-out of value added in France
Paul Cahu
May 2009
No. 58. Survey of household confidence and French consumer spending
Slim Dali
No. 57. Foreclosures in the United States and financial institutions losses
Stphane Sorbe
http://www.minefe.gouv.fr/directions_services/dgtpe/TRESOR_ECO/tresorecouk.htm
3.2.2 A more effective fiscal policy
As in Japan, Amer ica' s fiscal policy is heavily skewed towar ds
pr oviding shor t-ter m suppor t for final demand ( via tax
cr edits for households, pr ivate investment gr ants, public
investment, etc.) . Contr ar y to the Japanese exper ience,
however , the Amer ican stimulus plans may pr ove mor e effec-
tive. The $787 billion Amer ican stimulus plan, r epr esenting
5.5% of GDP and passed in Febr uar y 2009, coming on the
heels of the Febr uar y 2008 stimulus plan r epr esenting 1.2%
of GDP, consists of suppor t for consumption and help for
households in difficulty ( 39% of the plan, r epr esenting 2.1%
of GDP) ; 34% of the plan ( 1.9% of GDP) is intended for
investment in public infr astr uctur e investment, R&D ( in
science and ener gy) , and human capital; the r emaining 27%
( 1.5% of GDP) is mainly ear mar ked for help to States and
local gover nments, lar gely to fund Medicaid and Medicar e
pr ogr ammes
13
. Aid has also been made available for sector s
in difficulty, albeit on a r ather lesser scale; as a r esult, $55
billion ( 0.4% of GDP) was dr awn fr om the TARP
14
to come
to the aid of the car industr y.
3.2.3 Help for the financial sector
Over and above this suppor t for final demand, US monetar y
policy has displayed gr eat vigour in tackling the financial
cr isis and avoiding a Japanese-style deflationar y spir al.
In autumn 2008, the Tr easur y inter vened massively via the
TARP to r ecapitalise cer tain banks and nationalise the GSEs.
Subsequently, and in view of the inadequacy of the fir st set of
measur es, the new Tr easur y Secr etar y, Tim Geithner ,
announced a fr esh financial sector r escue plan mor e
massive, mor e ambitious, as well as better contr olled and
designed than its pr edecessor . The plan to buy up so-called
toxic assets was modified slightly ( see Box 5) . The banks'
defeasance str uctur e and the r ecapitalisation oper ations that
have taken place ar e compar able to the Japanese measur es,
but they wer e implemented far mor e r apidly.
To avoid the onset of a debt deflation spir al, the Amer ican
author ities have put in place a debt r estr uctur ing plan for
households in difficulty, in or der to stem the tide of mor tgage
defaults leading to depr eciation of asset values.
Sophie RIVAUD,
Michal SICSIC
(13) Medicaid is a programme that provides sickness insurance to low-income individuals and families. Medicare is also a
healthcare insurance programme, for individuals aged over 65.
(14) The TARP (Troubled Asset Relief Program) is a $700 billion programme created originally to purchase "toxic assets"
from the banks.
Box 5: TimGeithner's measures to tackle the financial crisis
$1,000 billion in public-private investment funds to buy up "toxic" assets;
Establishment of a stress test to assess financial institutions' capacity to withstand further losses and their lending capacity in the
event of a worsening of the crisis;
The bank supervisors, the SEC and the Treasury will persuade the banks to reveal their financial condition (i.e. their losses) in order to
improve market transparency;
A more realistic and forward-looking evaluation of financial institutions' balance sheets (Fed, FDIC, OCC and OTS).