The EconomistFebruary 1st 2020 Finance & economics 652
L
ike manygood intentions, such as
losing weight, the European Central
Bank’s inflation target is largely aspira-
tional. In 2003 it set itself a goal of below
but close to 2%, but in the past five years
inflation has averaged only 1%. That is
partly why Christine Lagarde, the bank’s
boss, began a review of the ecb’s mone-
tary-policy strategy on January 23rd. The
bank will ponder whether its target
needs adjusting or its tools sharpening—
and whether inflation figures capture
prices accurately in the first place.
Consumer-priceindicesaremeantto
reflectthecostoftypicalbasketsofgoods
andservices.Theeuroarea’shavea big
omission.Theycapturerentspaidby
tenants,butnotthecostsofbuyingand
owningproperty—eventhoughtwo-
thirdsofpeopleinthezoneowntheir
homes.AsBenoîtCœuré,whountil
recentlysatontheecb’s board,pointed
out,thebank’schosenmeasure“captures
onlymarginallythelargestsinglelife-
timeexpenditureofhouseholds”.
Housingisbothaninvestibleassetand, as a source of shelter, a consumable
service. But isolating the cost of the
service is tricky. American statisticians
assume that homeowners rent their
homes from themselves, and count the
“imputed” rent in consumer prices. This
has a weight of 11.5% in the index fa-
voured by the Federal Reserve. Japanese
and Swiss inflation figures also include
owners’ housing costs. (So does headline
inflation in Britain, but not the Bank of
England’s target measure.)
In fact, euro-area statisticians do
calculatethecostofbuyingandowninga
home.Addingit topriceindicescould
raisemeasuredinflationby0.2-0.3per-
centagepoints,notesDavideOnegliaof
tsLombard,aninvestment-research
firm.Thatisnothingtosneezeatwhen
officialinflationisonly1.3%.
Buttheeuroarea’sversionhasits
problems.It doesnotfullystripoutland
prices(roughly,theasset-pricepart)
fromthoseofbuildings(theconsump-
tionbit).Andunliketimely,monthly
inflationfigures,it appearsonlyquarter-
lyandaftera longdelay.In 2018 theEuro-
peanCommissionandtheecbdecidedto
leaveit outofconsumerprices.
Thestrategyreviewhasrevivedthe
subject.MsLagardewantstoexplain
bettertheecb’s policiestothepublic
(who,surveyssuggest,thinkinflationis
higherthantheofficialfigure).OnJanu-
ary27thYvesMersch,anecbboardmem-
ber,urgedtheinclusionofhousingcosts.
Fixingthestatisticalproblemswould,
asMsLagardewarned,taketime.Incor-
poratinghousingcostswouldalsoraise
questionsaboutthebank’starget:nudg-
inguptheinflationmeasurewithout
alteringthetargettoowouldlooklike
tippingthescales.Asdietersknow,
there’sneveraneasyfix.A fuller figure
Inflation in the euro areaThe ecbconsiders counting owner-occupied housing in inflationHomeimprovements
Potentialcontributionofowner-occupiedhousing*
toeuro-areaconsumerprices,percentagepointsSource:TSLombard*Assumesa 9%weightforhousinginthe
all-itemsindex †Excludesenergy,
food,alcoholandtobacco0.40.20-0.2-0.4
19181716151413122011Core†All itemssagemightoffercongratulations on win-
ning a prize, requiring only a small contri-
bution to unlock it. Your identity might be
stolen to make you responsible for repay-
ing a loan disbursed to somebody else. An
impersonator might steal your mobile-
money account when registering a new
simcard. Or your account’s security code—
the pin—might be leaked by a mobile-
money agent (one of the millions of small
traders whose business includes turning
mobile money into hard cash).
The poor may be especially susceptible
to such scams. They are more likely to be
relatively new both to the online world and
to formal financial services. And they are
less likely to have smartphones with so-
phisticated security software.
It is not just money that can be stolen.
So too can all the other data stored on a
phone. Often, however, much has already
been given away freely by its owners. Many
“free” apps are in fact paid for in customer
data. In every country people gaily sign
away their rights to privacy by clicking con-
sent buttons without having understood or
even read what they are agreeing to.
Indeed, in financial services the implic-
it sale of personal data is seen as a way in
which the poor can be “levelled up”. With
low, unreliable incomes and few assets,
they have been unable to borrow formally.
Now lenders are learning to rely on other
data to make credit judgments, notably
payment records, such as from a mobile-
money account, which can show a history
of reliability.
But such data can also be abused. Preda-
tory lenders and vendors might learn when
a costly loan or product would be hard to
refuse. Or an algorithm might (by design or
accident) be biased against certain borrow-
ers because, say, of their race or creed.
cgapmakes three policy recommenda-
tions. The first is to accept that the “con-
sumer-consent” model is irretrievably bro-ken, and to put the onus for looking after
the data on the service-provider. The sec-
ond is to give consumers full legal rights
over their data, allowing them to view, cor-
rect and move them without charge. The
third is to appoint “privacy representa-
tives”, who, among other tasks, would
check algorithms for signs of bias.
Many countries, from America to India,
are looking at improving data-protection
regulation. But cgap’s suggestions seem
very ambitious. “Free” financial services,
such as those offered already by the Chi-
nese giants, Ant Financial and WeChat,
have an obvious appeal.It is often assumed that the poor are re-
laxed about surrendering some privacy in
return for access to borrowing and other
services. In fact, concern about privacy is
not a preserve of the rich. Research in India
and Kenya has shown that even very poor
borrowers would be willing to pay a higher
interest rate—or join a much slower
queue—for a loan that came with more
guarantees that the data provided to the
lender would be kept private. Persuading
the tech giants that improved data security
for the poor is in their business interests
might be the best hope. But that will take a
lot more work. 7Where’s the scam filter?