Wiley, The London School of Economics and Political Science and
The Suntory and Toyota International Centres for Economics and
Related Disciplines are collaborating with JSTOR to digitize,
preserve and extend access to Economica.http://www.jstor.orgThe
Suntory and Toyota International Centres for Economics and Related
DisciplinesMr. Keynes' Evidence for Over-Population Author(s):
William Beveridge Source:Economica, No. 10 (Feb., 1924), pp.
1-20Published by: on behalf of andWiley The London School of
Economics and Political Science The Suntory and Toyota
International Centres for Economics and Related DisciplinesStable
URL:http://www.jstor.org/stable/2547868Accessed: 31-07-2015 10:28
UTCYour use of the JSTOR archive indicates your acceptance of the
Terms & Conditions of Use, available at
http://www.jstor.org/page/info/about/policies/terms.jspJSTOR is a
not-for-profit service that helps scholars, researchers, and
students discover, use, and build upon a wide range of content in a
trusted digital archive. We use information technology and tools to
increase productivity and facilitate new forms of scholarship. For
more information about JSTOR, please contact [email protected]
content downloaded from 164.11.203.58 on Fri, 31 Jul 2015 10:28:02
UTCAll use subject to JSTOR Terms and ConditionsMr. Keynes'
Evidence for Over- Population By SIR WILLIAMBEVERIDGE IN anaddress
on" PopulationandUnemployment"deliveredto theEconomic
SectionoftheBritishAssociation last September, I
criticizedcertainstatementsmadebyMr.J.M.Keynes.My address
hasbeenreprinted,according tocustom,intheEconomic Journal, for
December,1923, and isthere followed bya" Reply" fromMr.
Keynes,givinghisevidenceforthestatementswhichI challenged.Before
examining this evidence, letme trytoclear the issue, byindicating
whatis and whatis notin dispute. The one point of direct
disagreement is as to whetherEurope was already beforethe war
reaching thelimitsofthepopulationthatit
couldhopetosupportwithoutloweringitsstandardsofliving.
This,inrelationtothemainproblem ofpopulation,isapointof historical
detail,and itismuch more important tomethantoMr.
Keynes.Ifhisviewbecorrect-thatinEuropetheeconomic tendencies were
already adverse and the population excessive before thewar-therecan
belittlehesitationinregarding thepopulation as stillmore in
excessafter war's destruction of our resources ; the burden of
proof certainly lies on anyone who maintains the contrary. If my
view be correct, that Mr. Keynes' reading of pre-war conditions
waswrong andthattherewasthennosignofover-population in
Europe,itisstillopentohimtoarguethatinthechanged con-
ditionsthepopulationisexcessiveandthatachecktofurther
increaseisurgentlyrequired;theburden ofproof,however,lies on him.
There are other pointswhich are pointsofdisagreement onlyin sofar
asMr. Keyneshascometodefinite conclusions uponthem, whileIplead
forinquiry andsuspendedjudgment.Themost important of these is the
definite and immediate advocacy of " birth control " as a means of
limiting population.Itis not my view that
mankindcanwiselycontinuetopropagateindefinitelywithout considering
theconsequences;Iexpressedthecontraryviewas
emphaticallyasIcouldatLiverpool.Buttheconsequences and re-actions
of " birth control " also require consideration. There are other
points on which I am glad to think that we are in
completeagreement.Amongthemaretheinterpretation ofour present
troubles as mainly due on theone hand tothebreak-down I1 This
content downloaded from 164.11.203.58 on Fri, 31 Jul 2015 10:28:02
UTCAll use subject to JSTOR Terms and Conditions2ECONOMICA
ofinternationaltradeandco-operation, and on theotherhandto
uncertaintyastopricelevels;appreciation ofthedependenceof Britain
on the prosperity of other countries and on world-wide trade for
the maintenance of itspresentnumbers anditsinfluence in the
world;andrecognition oftheneedfor discussing theproblem of
population,bothquantitativeandqualitative,withoutprejudice and
withoutfear. One more preliminary issue mustbecleared.Inmyaddress
at Liverpool, I treated Mr. Keynes' fears of over-population as
relating toEurope asawholeor totheworld asawhole.Thatmyinter-
pretation was justified will, I think, be evident to anyone who
reads the full and literal quotations of Mr. Keynes' words in my
address or, better still,studies themin their contextin Mr. Keynes'
own book. Thefamous secondchapter ofthatbookiscalled " Europe
before theWar ";inthesectiononpopulationtheonlycountriesdealt
withbynameareGermany, Austria-Hungary andRussia.Later
comesthestatement: " InthechaptersofthisbookIhavenot generally
hadinmindthesituationsontheproblems ofEngland. ' Europe'in my
narration must generally be interpreted to exclude theBritish
Isles."Later still,referring back tothe second chapter by name, Mr.
Keynesinterprets what he saidthere asanassertion of" thediminishing
response ofNaturetoanyfurther increase in thepopulationoftheworld."
In his present Replyto me, however, Mr. Keynes more than once makes
the point that he was talking only of the " industrial countries
ofWesternEurope,"andusesthistorebutmyargumentsand
evidence.Thisis,Ithink,arealshiftofpositiononhispart.
Howfaritaffectsthevalueofmystatisticsofagriculturalpro-
ductionandofprices willbeconsidered later.Itdoesnotaffect atall the
criticisms which I have to make on Mr. Keynes' evidence.
Thesewillbefoundtoapplywithequalforce,whetheritis
theindustrialcountriesofWesternEurope,orallEurope,or
alltheworld,whoseover-populationbeforethewarisunder discussion.
WiththisprefaceletmepasstoconsiderationofMr. Keynes'
Reply,thatistosay,hisdefenceofthestatementswhichI
criticizedatLiverpool.ThesestatementsarecontainedinMr.
Keynes'well-knownbookontheEconomicConsequencesofthe Peace;as I
citedthem fully atLiverpool, Ido not reproduce them here.They are,
as Ithen said, incidental tothe main argument of that book or, as
Mr. Keynes rightly says, they are obiterdicta.This
doesnotmakeitanythelessimportanttodetermine, ifwecan, whether
theyare true or false.Nor mustMr. Keynes quarrel with mefor
criticizing atlengthhisobiter dicta aboutpopulation.For
onething,henowrepeatsanddefendsthem.Foranother,his
bookhasbeenreadby-atamoderatecomputation-halfa
millionpeoplewhoneverreadaneconomicworkbeforeand probably will not
read oneagain.Ithas coloured general thought This content
downloaded from 164.11.203.58 on Fri, 31 Jul 2015 10:28:02 UTCAll
use subject to JSTOR Terms and
ConditionsMR.KEYNES'EVIDENCEFOROVER-POPULATION3
onthequestionofpopulationand,inmyview,hascolouredit wrongly.
Theevidence on whichMr. Keynesrelies toshowthedanger of
over-population, whether inEuropeorinWesternEurope,before
thewar,isfoundin a comparison oftheaveragepricesofBritish
exportsandimportsinsuccessiveyearsfromI873.Theoriginal authority
for these figures is Professor Bowley;from the comparison ofprices,
heprepared andpublished intheEconomic Journalfor
I903anindexnumberof" thequantityofexports(thekind
supposedunchanged)thatweregivenforauniformquantityof imports" in
each year from I873to I902.From I900export and import prices on a
somewhatdifferent basis havebeen prepared by the Board of Trade and
are used by Mr. Keynes to carry on Professor Bowley'sfigures.'
These figures were notunknown tome, butIconfess thatIhad
nottillnowconsidered themcarefully.Whenmyattentionwas
calledtothemwhileIwaspreparing myLiverpoolpaper,they seemed to me
to have so littlebearing on the question of population thatIdid
notinvestigatethemfurther.HadIrealized thatthey were thebasis ofMr.
Keynes' statementsIshould, ofcourse, have discussed them. The
actual figures cited by Mr. Keynes in his reply to me aretwo-
fold.One set, which may be described as the " general index," deals
with British exports and imports gelnerallyfrom I873toi9ii.The
other, or " special index," deals with selected groups of articles
only -manufacturedexportsandfoodimports-fromi88ionwards.
BetweenI88IandI902theonlyexportsincludedare " textiles" and "
metalproducts."Mr. Keynestakesthemeanoftheindex numbers ofprices
givenby ProfessorBowleyfor thesetwogroups
torepresentmanufacturedexportsasawhole;expressesthe
correspondingprice index for imported food, drink, and tobacco, as
a percentage of thismean, and so getsan indexwhich maybe des-
cribed as representing either the" volume of manufactured exports
givenforauniformquantityoffoodimports,"orthe " priceof food imports
as a percentage of the price of manufactured exports." From
I902onwards he uses theBoardofTrade prices for " articles whollyor
mainlymanufactured " andfor " importedfood,drink, and tobacco."Itis
tothespecial index which he constructs in this waythatMr.
Keynesattachesmostimportance; for convenience Ireproduce here
thefigures givenbyhIiimwithten-yearaverages, upto9I3.2 1
ProfessorBowley'smemorandaareintheStatisticalJotirnalforI897 (pp.
437-9)andtheEconomicJournalfor1903(pp.628-32).TheBoardof
Tradefiguresfor1900-IIwerepublishedinanofficialpaper,Cd.6314,andare
continuedupto1913inCd.6782andCd.7432.Mr.Keyneshasanoteon
themonp.630oftheEconomicJourncal for19I2. 2
Mr.Keynes,forsomereason,carriesonhistableto i9i6. Theselater
figureshavenobearingonthequestionofwhatwashappeningbefore thewax.
This content downloaded from 164.11.203.58 on Fri, 31 Jul 2015
10:28:02 UTCAll use subject to JSTOR Terms and Conditions4ECONOMICA
VOLUMEOFMANUFACTUREDEXPORTSGIVENFORAUNIFORM
QUANTITYOFFOODIMPORTS(I902=IOO).
YearlyTen-YearYearlyTen-YearYearlyTen-Year
Figure.Average.1Figure.Average.Figure.Average. i88II32 I89Iio6 II7
I9OI95 I05 I882130 I892II2 115 I902100 104 I883I28I893II2 II3I903
IOO I03 I884I20I894I07 II2 I904 96I02 I885 II4 i895104 III I905
99IOI I886 II7 I896io6IIOI90692100 I887ii6I897IIOlO9 I907 9298 i888
II4 I898IIIIO9I9089897 I889II6I899 I05 io8I909 I04 97 I890 I07
II9I90093 I07 I9IOIOO98 I9II9898 I9I2I0298 I9I39798 Commenting
ontheten-yearaverages, Mr. Keynesargues that down tothedecade
ending I907-8they show an "uninterrupted iin- provement,"
thatis,asteadydeclineinthequantityofexported manufactures required
tobeexchanged for auniform- quantityof imported food, and that with
thisdecade, ofwhich themiddle year is I903,the improvement came to
an end.Inthelight of this table, his "formergeneralization should
be modified by placing the turning- point atI903,rather than
atI900(the figuresof the period I899- I903beingmuchupsetforpurposes
ofcomparison bytheSouth African
Warandbytheboomandsubsequentdepression within thatperiod)
andbyemphasizing moredecidedlythecessationof improvement and
lessdecidedly theextentof theactualdeteriora-
tionbetweenthatdateandtheoutbreakofthewar."Thisis
substantiallythewhole ofMr. Keynes' case. The first comment that
must occur to any impartial reader is that evenas givenbyMr.
Keynes,thefigures are singularly indecisive. The index falls from
II9in the decade I88I-I890to 97 in each of the decades
I899-I908andI900-I909; itthenrises onepointandre- mainssteady at 98
for four years, tothe decade I904-19I3. Even onthesefigures
theturning-point, such asitis,comesinI904or I905,ratherthaninI903;
theactualminimumforindividual years was reached in I906and I907.But
is it sensible to speak ofa " turningpoint "atall
?Anindexwhichhasfallenpointby pointfor twentyyears,remains
steadyduring atimeaffected(as Mr. Keynes himself points out) by
many special circumstances-the SouthAfrican War,
theexceptionaldepression ofI908-9,andthe upward
turnofwholesaleprices. Thefigures themselvesgive no reason for
assuming that,had no war come in I9I4,theindex would have remained
stationary or haverisen; it might just as well have
beguntofallagain.Thegeneral indexwasactuallyfalling each 1 The
ten-year averages are entered opposite the last year included. This
content downloaded from 164.11.203.58 on Fri, 31 Jul 2015 10:28:02
UTCAll use subject to JSTOR Terms and
ConditionsMR.KEYNES'EVIDENCEFOROVER-POPULATION5
yearfromI9I0toI9I3.1Toinferaturning-pointinthehistoryof European
civilizationfromamovementofthisnaturewouldbe rash beyond
justification, even if the figures themselves were correct, if
theyapplied tothe whole ofEurope or Western Europeandnot merely
topart of thetrade of one country,if no similar movement had
occurred in them before, and if, finally, they had any bearing on
thequestionofpopulationatall.Actuallynotoneofthesefour conditions
is satisfied.The figuresas given are statisticallyfallacious and as
corrected show no turning-point atall;theyapplytopart of British
trade only,and corresponding figures for Germany show
nocorresponding movement;ifcarriedbackbeforei88Ithey show an upward
movement then far more marked than Mr. Keynes finds later;finally,
they are in principle irrelevant to the issue.Let me
justifythesecriticisms one byone. I.-Thefigures are fallacious and
if correctedshow no turning-point at all.Mr. Keynes, in
constructing his index, uses figures from two sources-fromProfessor
Bowley for the yearsi88itoI902,in which the only exports included
are " textiles"and " metal products," and
fromtheBoardofTradefortheyearsI902toI9I3,inwhichall
exportedmanufacturesareincludedforwhichquantitiescanbe given;these
two sets of figures he treats as equivalent and combines
intoasingle series.Unfortunatelythefigures are notequivalent. If
Mr.Keynes, instead of using the Board of Trade figures after I902,
when Professor Bowley's figures came to an end, had himself carried
on the latter, calculating them on Professor Bowley's lines, he
would have reached quitedifferent results.Hisindex numbers and ten-
yearaverages wouldhavebeenasfollows:
Inidexforyear.Teni-YearAverage. 1902.........100 104 1903*.-... ...
97103 1904.........93I01 1905......94 100 I 906.........9 1 99
1907.........9097 1908.........9395 1909.........9995
I9I0.........9495 19II.........9094 1912.........9494 1913... .....
879 3 The result isstartling;thedecline from i88itoI903isseencon-
tinuingwithoutinterruptionuptotheeveofwar.Theactual evidence
submittedbyMr. Keynes inhis Replyisthusseen to be
worthless;hehascombined twosetsoffigures whosemovements do not
agree and he getsa turning-point near I900onlybecause at thatpoint
he himself turns from one setof figures toanother. IThis
waspointedoutintheBoardofTradeReturnpublished in 1914.
"Onthewholeitwouldappear that during recent yearstheaverage values
ofBritishexportshavebeenincreasingmorerapidlythanthoseofimported
rawmaterialsandfood." This content downloaded from 164.11.203.58 on
Fri, 31 Jul 2015 10:28:02 UTCAll use subject to JSTOR Terms and
Conditions6ECONOMICA Thisdoesnotsettlethequestion.Themarked
discrepancy be- tweenthecourseoftheBoardofTradefigures
andofProfessor
Bowley'sfigures,ifcontinuedafterI902,suggestedtomethe
desirabilityofacloser examinationofboth;examinationshowed
thenecessityofrevisingProfessor Bowley'sfigures.The" metal products
" covered byhis statisticsare named by him as " iron and
steelandmanufacturesthereof,copper(wrought),brassofall sorts " ; in
estimating their average price, the first of these headings, "
ironandsteelandmanufacturesthereof,"hasbeenlumped
together.Thisisunsatisfactory,becausetheheadingcoversa
greatvarietyofarticlesofdiffering degrees offinish anddiffering
price-frompig-iron and rails to tin-plates,wire, and wrought tubes
andpipes.Theaverage valueper tonofiron andsteelmanufac-
turesasawholemaychange,notthroughareal change ofprices
(i.e.,ofthesumgivenfor thesamearticle atdifferent times),but
because the composition of those manufactures has changed.There
has, in fact, been a marked and progressive change in the character
ofthosemanufacturessincei88i;themorefinishedandmore costlyitemsform
agrowing proportion ofthewholeandraiseits average price bychanging
itscomposition.This error can onlybe avoidedor reduced bya detailed
classification of the manufactures
andbycalculatingthechangeofpriceseparatelyforeachsub-
division,asisdonebytheBoardofTradeafterI900.Ihave
accordinglycalculatedafreshthepriceof" metalproducts"as defined
byProfessor Bowley,splittingupiron andsteelmanufac-
turesintotheseventeensub-divisionsforwhichseparatefigures are
givenbacktoi88I.The price oftextileshasalsobeenrecal-
culatedwithacloserclassification;thechangeinthefigures is much
smaller than withmetalproducts, butinthesame direction.
Norecalculation oftheprice offoodimports wascalled for.All three
series of prices-exportedtextiles, exported metal products, im-
ported food-havebeen carried backtoI865.The revised figures are
setoutin Table I.Theresultoftherevisionon Mr. Keynes' position is
made clear bythe accompanying chart.(p. 8.)
Thebrokenlinebeginningini88i(TextileandMetalExports Unrevised) is
basedonProfessor Bowley'sfigures, as givenbyhim
uptoI902andcontinuedonthesame principle toI9I3,i.e.,with
ironandsteelmanufacturestreatedasaunit.Thedottedline, beginning in
I902(Manufactured Exports) is basedon theBoard of Trade prices of
exports " whollyor mainlymanufactured'" and of
"importedfood,drink,andtobacco."Thefiguresgivenin Mr.
Keynes'table(p. 4above)are representedbythebroken line
toI902andbythedottedlinethereafter.The"turning-point" appears
becauseMr.Keynesswitchesfromonelinetoanother; evenso itis
barelyvisibleand whollyindecisive. The continuous line from
i865toI9I3(Textile and Metal Exports
Revised)isbasedonmyrecalculationoftextileandmetalprices direct from
theStatisticalAbstracts toI899and from theBoard of This content
downloaded from 164.11.203.58 on Fri, 31 Jul 2015 10:28:02 UTCAll
use subject to JSTOR Terms and
ConditionsTABLEI.-BRITISHEXPORTANDIMPORTPRICES (1902=100).
ManufacturedExports ~~FoodImportPrice ManufactulredExports as
percentageof Year. Food Manufactured Export Textile,Metal.Mean
ofImports, Price. I and 2.10 year Yearly.Average. I. 2.3.4-5.6.
I865189132I6II38 86 - I866 193 135I6414286 - I867177128153 156102
i868I66123145155107 i869167121144145 101- I870 I59124141138 98 -
JI87I 157I22140I48 io6- 1872I65I68I66 I53 92 - I873 I58192175 15488
- I874 149176I62I5394 9 I875 146r15 148 1I469897 I876
135138137144I0599 I877 132117 124I55I25I01 I878128114I21 I40II6 102
I879123104 113I33I17104 I88o 125II6121 139115 io6
I88I121105113I39123107 I882124Io8II6I39I20 110 I883I22102112 133
II9 II3 I884ii896107I22 114115 I885 11389IOI112III 117 i886io8 85
96io8 I12 117 I887io882 95 io6112ii6 I888io88697 io8112II6
I889I0890 99 III112 115 I890Io61OI104I08 104114 I891 I0599 102 114
III 113 I892100 9497 110 114 II2 I893IOI 8995 I06II2 III I89496 85
9I96Io6III I89593838893 io6IIo 1896978491 95105 1O9 I89795869098
lO9lO9 I8989287go99 110 1O9 I8999399 9698102 I08 i900 103II8 110100
9I I07 I9OI 103 I04I03 10198 I05 19021OO100 1OO 100100 104 1903IOI
101101100 99103 1904110 95 102 9997 102 1905III 95103 101 98IOI
I906iI8100 I09 101 93 100 1907125Io8II6Io69I 98 1908I20 101IIIio898
97 I90911597 io6III 105 97 I9IO I2494IO9I12 I03 98 I91-I2995I12 III
9998 1912127 100 II3117103 99 1913 I32 Io6II9 II5 97 98 This
content downloaded from 164.11.203.58 on Fri, 31 Jul 2015 10:28:02
UTCAll use subject to JSTOR Terms and Conditions8ECONOMICA
I.-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Q~~~~~~~ '4j o- I- q s
This content downloaded from 164.11.203.58 on Fri, 31 Jul 2015
10:28:02 UTCAll use subject to JSTOR Terms and
ConditionsMR.KEYNES'EVIDENCEFOROVER-POPULATION 9 Trade Return
fromI900toI9I3.Itisasreasonably accurateas
sucharecordcanbemade,andshouldbesubstitutedforthe figures used
byMr. Keynes.Whatdoes itshow ? Inthefirst place, therevised
indexcrosses theincorrectone;it is materially below
thelatteratthebeginning oftheir jointcourse (I88I)and as much above
it at the end (I9I3).The result of revision istomake theprice of
bothtextilesand metals higher ini88iand lower
inI9I3thanitwasintheunrevised figures, becausethein- fluence of the
change in composition, i.e.,of the growing proportion
ofmorefinished andmoreexpensivearticles, iseliminated.The
volumeoftextileandmetalexportsexchangedforauniform quantityoffood
imports falls less rapidly in truth thanitappeared todo
intheunrevisedfigures.
Inthenextplace,thereisnothingthatcanbedescribedasa turning-point
inthegeneral trendofthecurveatoraboutI900. Itistrue thatifwe follow
Mr. Keynesintakingten-yearaverages oftherevised figures as ofthe
old we getoncemore a fall,though lessmarked,frorn
I88I-90toI900-09followedbyatriflingrise. Butthisonlyillustrates
theimperfection ofamechanical ten-year average.Whatthecurve really
shows is afairly steady downward slopethroughoutitscourse
fromi88itoI9I3,exceptwhere itis broken
byanexceptionaldipbetweenI900andI907.Afterthis dipitrises
toalevelrepresenting analmostexactcontinuationof itsformer
course;thepeaksatI877,i88i,I892,I899,andI909
areroughlyinaline.Thediprecords obviouslytheworkingof special and
temporary causes of which theSouth African War was thegreatest.Our
newcurveagrees withmuchotherevidencein picking outtheEdwardian
ageasamomentary exceptionalphase inBritish economic
history.Butitisas far aspossible from sug- gestingasecular change
nearI900. ThecorrectionofthefigurestakenfromProfessorBowley's
memorandum really disposes ofMr. Keynes'case.Inviewofthe importance
of the issue, however, itis worth whiletoproceed with
theothercriticismsofit.EveniftheindexconstructedbyMr. Keyneswere
thought to show a change of secular trend aboutI900 or should do so
later, this would be no evidence of over-population. II.-Thefigures
relate to a part ofBritish trade alone and are not typical of
othercountries such asGermany.Icriticized Mr. Keynes atLiverpoolnot
for anything he had said about Britain, but for what he had said
about Europe, and, as Ithought,thewhole of Europe.
InhisReply,henowtreatshisargument asreferring only tothe "
industrial countries ofWestern Europe? " Butistheparticular British
experience on which he bases himself-therelative movement of
certain export and import prices-typicaleven of other industrial
countries of Western Europe?I have testedMr. Keynes' argument
forthemostimportant ofthesecountries-Germany-byworking out series
of figurescorrespondingas closely as possible to the British
ones.The results are shown in Table II,givingfromi88itoI9II This
content downloaded from 164.11.203.58 on Fri, 31 Jul 2015 10:28:02
UTCAll use subject to JSTOR Terms and Conditions101NECONOMICA
TABLEII.-GERMANEXPORTANDIMPORTPRICES(1902=100).
ManufacturedExports.Mean ofImports of British Food
Year.AllTextilesFood, Drink,Import Price
Textile.Metal.Other.Manu-and Metalsand Tobacco.as percentage
.factures.of column(4). ()(2)(3)(4)(5)(6)(7) I88i12388125120 105144
II6 I882 113157124121 135136115 I883112 144132123128 134 Io8
I884Io8138114114123II9 107 18851031OO112107102 lOg 105 I886Io693112
107 100110101 I8871059811210710211799 I888 105105 112io8105 II4 100
I889110 114115113 112I2698 1890lO9 107 II6IIIIo812798
I89I9996io810297 141 112 189292871009589125II6 I893926898918o117117
18948585948985 IOIIo8 18958986969287102101 1896889398939099102
1897887896939393105 189889100929395101 107 1899951129798 104 97100
1900 10311398102Io810298 I9O1981001011OO99102100
1902100100100100100100TOO 1903io693981OO1OO95100 1904io8 85
99100969999 1905114861031041OO10797 190610482 107 10193 105 100
1907Io68311210494113102 I9081O976III10292112106
I90991741079683II61i6 I9IO937611098 85105114 19I19587 11010191 117
110 119112 1913 I914
priceindexnumbersfortextile,metalandothermanufactured exports and
for all these together, and for food imports.For reasons described
in the Appendix the last of these indices is not a good one;
inthelastcolumn ofTable IL, therefore, Ihavesetoutaseries of
figures representing therelativepriceofGerman exportedmanu-
factures and of British food imports, taking thelatteras typicalof
the supplies sent by the new world tosustain the old.These figures,
itwill be seen, show no sort of trend, butsimply a large
fluctuation betweenthesame upper and lower limits-ii6or II7for
thecrests This content downloaded from 164.11.203.58 on Fri, 31 Jul
2015 10:28:02 UTCAll use subject to JSTOR Terms and
ConditionsMR.KEYNES'EVIDENCEFOROVER-POPULATIONll and 97 or 98 for
the depressions.If we substitutetheGerman food importsforBritish,
or useanyofthe otherGerman figures, we get curvesequallyorevenmore
unlike theBritishone.Mr. Keynes' assumptionthathisBritishfigures
aretypicalofindustrialcom- munities generallyiscontrary tofacts.'
Thisisnotsurprising whenweconsider whathisfigures really
represent.Theyrelate noteventothewhole ofBritishtrade but only to a
part of it.For " manufactured exports," Mr. Keynes uses
uptoI902themean of separate figures givenbyProf.Bowleyfor "
exportedtextilemanufactures " and'exportedmetalmanu- factures."The
former does cover mostof our textiletrades.The lattercovers
onlywhat isclassified in thetrade returns as " manu- factures of
iron and steel,copper (wrought) and brass;thatis to say, it
excludes machinery, railway and other vehicles, ships, cutlery and
hardware, andelectrical engineering.Allmanufacturesother
thantextileand metal,such as dress, chemicals, wood, paper, soap
andfurniturearealsoexcluded.Invalue,Prof.Bowley'stwo classes of
textileand metal manufactures represented ini88iabout J
35,800,000;this was 65 per cent. of all our exported manufactures
and 58 per cent.ofallour exports.Fori895,thecorresponding
figureswere?II4,500,000,59percent.and5Ipercent.2This
evidencethuscoversafallingproportionofourtotalexportof
manufactures:itomits,perforce, from " metals"themostskilled
andmostprogressive sections;itrepresents atbottomlittlemore than
cotton,wool, tinplates,rails, galvanized sheets and castings.
FromI900onwardsthebasisofthefigures isdifferentandin some ways
wider.But the widening of the basis is to a large extent
illusory;neither before nor after I900is itpossible tomake certain
thattheships,machinery,motor-cars andothercomplicatedand
infinitelyvariousmanufacturescountedorweighedeachyear represent in
each year comparable articles.The most characteristic and most
progressive section of British industry is the leastsuscep-
tibleofstatisticalmeasurement.3 Some British figures, no doubt, can
be taken withoutserious risk, as broadly typical of conditions in
Europe or in the world as a whole. In my Liverpool paper, with due
warning, I used the Sauerbeck and 1
ProfessorBowleygavesomefiguresforGermanyintheEconomicJournal
forI903.Thesedealtwithexportsandimportsgenerally,butreferredonly
toselectedyearsofwhichI900wasthelatest.Theyindicatedamovement
similarindirectiontothatofBritain,butmuchsmallerinamountandwith
markeddifferencesinparticularyears.Theycertainlydidnotsupportthe
viewthattheBritishindexnumbercouldbetakenasaccuratelyrepresenting
Germanconditionsaswell.
2ForI9IItheywereabout?204,000,000,56percent.and45percent. 3
Fromi88itoI9IIexportsoftextilesandofironandsteelandtheir
manufactureseachroseinvaluejustover6opercent.Exportsofmachinery
rosemorethan200per cent.,andthoseof
chemicalsonlyslightlyless;apparel
roseI50percent.;railwaycarriagesandothervehiclesrosefromp227,000 to
?8,I25,,ooo; AlltheseareexcludedfromthefiguresusedbyMr.Keynes
toI902. This content downloaded from 164.11.203.58 on Fri, 31 Jul
2015 10:28:02 UTCAll use subject to JSTOR Terms and
Conditions12ECONOMICA Board of Trade indices of wholesale prices
inthis way, to bring out certainstrikingdifferences
betweencoal,cornandotherarticles. ButMr.
Keynesoughtnottohaveassumed, withoutinquiry,that thelimitedfigures
ofBritishtradegivenaboveareatalltypical
ofthetradeofothercountries;stilllessthattheyare soexactly
typicalthattheslightandindefinite movementobserved in them must
have occurred elsewhere.Mr. Keynes' index is vitallyaffected
bythepeculiarcompositionofBritishtrade,bychangesinthat
composition,bycyclicalfluctuation,and,inthecriticalopening
yearsofthiscentury, bytheSouthAfrican War andbyahostof
specialcircumstances. III.-TheBritishindexhadrisenpreviously wlten
there wasno over-population.Evenifthere hadbeenamarked
riseafterI900 in the volume of manufactured exports given for a
uniform quantity of food, there would havebeen nothing exceptional
in this.Let us returntotheBritishfigures inTableI,andinthechart.Mr.
Keynesbegins hisfigures justatthepoint(i88i)from which they happen
toshow a steadyfall.If he had gone a littlefurther back, he would
havefound a yetmore striking rise.'The ten-year aver- ages mount
steadilyfrom 96 ini865-74toII7iR1877-86,a move- ment which makes
the subsequent rise from 97 to 99 or 98 just before
thewarlookinsignificant-as,indeed, itis.Ariseintheindex
obviouslycannotbyitselfbetakenas evidenceofover-population or
diminishing returns.Whatisthereal lessonoftheinteresting
andstrikingmovementsthatitshows ? The first lesson is that we
should be cautious in drawing ainylesson
atall.Thisissuggestedbytheobserved serious riskoferror in treating
any large class of exports(such as iron and steelmanufac- tures) in
the lump, and by the need for the most detailed classification
possible, iferror istobeavoided.Professor Marshall ina striking
passageoncecalledattentiontothecontinualmovementtowards
themoreadvancedstagesofmanufacturing,tothefinishingand final
processes, asanoutstandingfeatureofBritisheconomic life, and
asthesymbolofitsindustrial rank. " One testofleadership
isdoingthingswhichothercountries withsimilar economicprob- lems
will be doing a littlelater, but are not ready and able to do yet.
One ofthebestindications ofthenature and extentofacountry's
leadershipistobefoundinthecharacterofthegoodswhich
sheexports."2Suchleadershipisillustratedateverypointof our
exporttradeandprices.Almostwithoutexception,inall our main
exporttrades, we find themore finished and morecostly growing
attheexpenseofthelesscostlyor atleastmore rapidly;
whenever,inthetradereturns,twoormorearticlespreviously 1 The "
GeneralIndex " givenbyMr.KeynesfromI873,butnotreferred
tobyhim,showstheendofthisrise. 2
IndustryandTrade,p.3.Comparealsoanotherpassageonp.26: " In-
dustrialleadership...isthepowerofdoingnowwhatotherswillpresently
bedoing,oratalleventstryingtodo." This content downloaded from
164.11.203.58 on Fri, 31 Jul 2015 10:28:02 UTCAll use subject to
JSTOR Terms and ConditionsMR.KEYNES'EVIDENCEFOROVER-POPULATIONI 3
combined are separated, the more costlyor more finished ofthem is
seen thereafter togainontheothers.Tinplates,wire, galvanized sheets
gainonpigiron,bolts,bars,andrails;wroughtoncast
iron;steelmanufactures on steelingots;bleachedyarnon grey;
printedanddyedpiece goodson plain ones;all-woolclothsand
coatingsonmixtures.Thisisafeatureofoureconomiclifeen-
couragingpractically,butratherdepressingtothestatistician. Every
improvement in the classification of our trade returns, making
itmore nearly certain thatweare comparing likewithlike,makes
theaverage price in later years appear lower thanit would without
thatimprovement.Butnoclassificationcanbeperfect.We cannot,or,
atleast,donotpublishseparatestatisticsforeach different qualityof
yarn or piece goods,or class of pipesor wire or girders;wecan
onlyinour indexofprices allow for thechanging character of British
trade when the change shows itself in the relative
growthofseparatecategories;achangewithinacategorygoes
unmarked.Yetsuchchangestherecertainlyare;itiscommon knowledge
thatonthewholethetendencyofcottonspinning has beentowards the finer
counts.Ultimatelyitisnotpossibletosay
howmuchoftheriseofexportpricesrelativetoimportprices
showninthefallofourindexfromi88iisapparentonly,how much represents
a real gaininpurchasing power ofmanufacturing intermsofagricultural
products.One canionly say thatthe main movements of the index,
bothup beforeI877anddown afteri88i, lookfar toolarge tobe
explainedawayaltogether. Withthiswarning,
wemayseecrystallizedinthemovementsof our index,a good deal of
interesting economic history.Theindex represents a fraction, with
the price of food imports in the numerator and theprice
oftextileand metalexportsinthedenominator.It rises whenthelatter
become relativelycheapandfallswhenthey
becomerelativelydear;inparticular itfallswithgoodharvests and as
the after-effect of wars.The American Civil War, the Franco- German
war, and the South African War all raised the relative price of
manufactures.We are at this moment repeating this experience; the
index for i92i,I922and the first nine monthsof I923has stood at73,
77, 8o.1The mostmarked feature of thecurve, however, is the sheer
rise from I873(the boom afterthe Franco-German War) to I877and the
long downward trendthereafter.The natural inter- pretation is that
in the early 'seventies we were shipping tothe new lands beyond the
seas large quantities of railway material and similar products, and
thereafter reaped theharvestinagrowing stream of cheaper
food.2Nowthecurvehascomebacktosomethinglike the levelseen before
theuprush ofthe'seventies. One thing isclear-asharp rise
oftheindexmay herald notthe 'SeeMr. Keynes'Reply(pp.
480-IoftheEconomicJournal)for this,tohim, ratherdisconcertingfact.
2 Thevolumeof railwaymaterialexportedin thefouryearsI869-72wasnot
equalledagaintillI887-90. This content downloaded from
164.11.203.58 on Fri, 31 Jul 2015 10:28:02 UTCAll use subject to
JSTOR Terms and Conditions14ECONOMICA
end,butthebeginningofaneraofprosperity.Theremustbe somethingwrong
withthelogicofanyargumentthatconnectsit inevitablywith the
former.This brings me to my final criticism. IV.-Arise in the index
has no logical connectionwith over-popula- tion.Evenif the index
used byMr. Keynes were representative of thewholetradeofEurope
andnotmerelyofpartofthetradeof Britain, and even if its movements
since I903showed anunmistak-
ableandunprecedentedupwardtendency,thiswouldaffordno
evidenceofover-population or of adiminishing response of Nature
toman's efforts.There is a gap in the argument which twoquota-
tionsfromMr. Keynes'writingswillmakeclear. The first quotationis
from his noteon the Board of Trade statis- ticsof export and import
prices, in theEconomicJournal." There
isnowagain,"hetherewroteini9i2,"asteady tendency fora
givenunitofmanufactured product topurchaseyearbyyeara diminishing
quantityof raw product.The comparative advantage
intradeismovingsharplyagainstindustrial countries."
Thesecondquotationisoneof thosealreadygivenbymeat
LiverpoolandrepeatedbyMr. KeynesinhisReply.'" Upto
aboutI900aunitoflabou-rappliedtoindusiryieldedyearby year a
purchasing power over an increasing quantityof food.Itis possible
that about theyear i90othis process begantobereversed
andadiminishing yieldofnaturetoman'seffortwasbeginning tore-assert
itself."The italics in each case are mine.
Thefirstquotation,apartfromtheunjustifiedinferencefrom Britain to
all industrial countries and the exaggeration of the words "
movingsharply,"isadefensiblecommentonthefiguresthen under
review.Thesecond quotation,whichMr. Keynestreatsas no more than a
translation of the first, is something absolutely differ- ent. A "
unit of manufactured product " and a " unit of labour applied to
industry " are no more identical terms than are a pair of boots and
a boot-maker;thetransition from one totheother is indefensible. Mr.
Keynesignoresallchangesinmanufacturing technique,both
absolutelyandrelativelytoagricultural technique. The price of any
product is vitallyaffected bythe cost of produc- tion.The main
objectand usual effect of improvements in manu- facturing technique
are to lessen the costs of production, to produce
alargerquantityofgoodsbythesameamountoflabour,and (almost
inevitably) to lower the price in order to find a market for the
largerproduction.Amarkedimprovementinthetechniqueof making
anyparticular product ofgeneral utilityalmostinevitably lowers
theprice relativelytootherthings,thatistosay,thepur- chasing power
ofeachunitofthatproduct.Butitincreases, not diminishes, thereturn
toeach unitof labour applied tomaking the product.Thephenomenon
whichMr. Keynestreatsasevidence Qf over-population and
ofdiminishing returns, namely,decrease in 1
FromTheEconomicConsequences ofthe Peace. This content downloaded
from 164.11.203.58 on Fri, 31 Jul 2015 10:28:02 UTCAll use subject
to JSTOR Terms and ConditionsMR.KEYNES'EVIDENCEFOROVER-POPULATIONI5
theprice ofBritishmanufactures relativelytofood,would cccur, other
things being equal, whenever industrial technique was advanc-
ingmore rapidlythanagricultural technique;itmightoccur,in
suchcircumstances, whetherpopulationwasgrowing orfallingor
stationary, whether returns to labour in agriculturewere increasing
or diminishing.The special index is simply an expression of the
price of certain manufactures in relation to the price of food.The
course ofsuchanindexistheresultantofseveralindependentforces,
namely,efficiency ofproduction inindustryorinagriculture and
demandforindustrialoragriculturalproducts.Herearefour variables at
least.There is no justification for attributing a change in the
resultant to one of these variables alone, and building thereon
asweepinggeneralizationofdiminishingreturnsinagriculture. The
change observed is just as good evidence of increasing returns in
industry.Thequestionofmanufacturing efficiency isdealtwith byMr.
Keynesin hisreply as follows:" Fromi88itoi9oothere
wascertainlyagreatimprovementinmanufacturingtechnique. Thus,
notonlydid a unitofmanufacture purchase more food,but itcost less
labour....Between thattime(I900)and the outbreak
ofthewarthesituationmayhavebeenhelpedbysomefurther improvement
inmanufacturing technique;andthevolumeofour
tradewasstillincreasing;buttheotherfactor-thepurchasing power of
our manufactures-ha.dceased to help and seems from the
abovedatatohavebeguntoturnadverse."Hethenmaintains thatthe "
bestevidence available " stillsupports his contentionas
tothefallingreturn sinceI900toeachunitoflabourappliedto
industry.From this passage Iinfer thatMr. Keynes has begun to
suspect the gap in his argument and is making a half-hearted
attempt tofillitbysuggestingthatimprovementsoftechniquesomehow
ceased or become negligible after I900; that this variable,
therefore, maybe leftoutofaccountandthewholechange beattributedto
diminishing returns inagriculture. Iknow of no grounds for
believingthatimprovements inmanu-
facturingtechnique-includinginvention,replacementofoldby
newplant,andprogress intechnical education-becameless rapid
afteri90othanbefore.YetforMr. Keynes'case,itisnecessary to hold not
merely that progress became less rapid but that it stopped
altogether;only on thatsuppositioncanitbeassumed thateach
unitofmnanufacturedproductrepresentsthroughoutthesame arnount of
labour applied toindustry. AllthroughMr.
Keynes'Reply,thereisnorecognitionofthe effect of technical
improvements in cheapening the product, thatis to say, of the
connection between cost of production and price.The point was dealt
with, however, in myLiverpool paper, in a passage to which Mr.
Keynes has really not paid enough attention.I there
explainedthatoneway,insubstancetheonlyway,inwhichthe rising price
ofcoal and metalintermsofcorn could bereconciled with a falling
price Qfindustrial products based on coal and metal in This content
downloaded from 164.11.203.58 on Fri, 31 Jul 2015 10:28:02 UTCAll
use subject to JSTOR Terms and Conditionsx6ECONOMICA terms of corn,
was through an increase of manufacturing efficiency.
Ipointedoutthatthiswasnotonlyareal possibilitybutaprob-
ability."Itisjustinthespheres ofmanufacturing and distribu-
tionthatincreased efficiency mostnaturallyaccompanies agrowth
inpopulation,andthatinventionandorganization wintheirlast
victoriesoverdiminishing returns.Butthecheapening ofmanu- facture
in this way involves not a decreasing but an increasing return
toeach unitof labour in industry;itwould cause a fall of thereal
cost of corn measured in labour."Actually,the combination of Mr.
Keynes'figures, of therelative prices of exported manufacture and
importedfood,withmyfigures oftherelativepricesofcoaland
metalascompared withcorn,suggestthattherewasprobablya
substantialimprovement of manufacturing efficiency betweeni9oo and
theoutbreak of thewar.Ifwefirst and foremost consider, as from the
point of view of this country we are justified in considering,
labour in manufacturing as distinct from labour in extractiveindus-
tries, itis all but certain thata unit of labour applied to
industry in thiscountry continued after I900as before to secure
year by year a purchasing power over a continually increasing
quantityof food. ThetablewhichIhavebeencriticizing represents
thewholeof Mr.Keynes'evidence.Atleast,hehascallednootherwitness.
Under cross-examination thewitness breaks down hopelessly. His
character is bad;he doesn't say what Mr. Keynes wants him to say;
andwhateverhesaidwouldbebesidethepoint.Hewasnever within a hundred
miles of the murder that he is asked todescribe.
Ontheotherhand,thereasonsthatIgaveatLiverpool, for doubting whether
there had been any murder at all, are untouched by Mr. Keynes'
criticism.Of theweakness of wholesale prices, limited
practicallytorawmaterials,asproofoftheratioofexchange
betweenindustrialandagriculturalproducts,Iwaswellaware;
ingivingwholesalepricesIdealtatsomelengthwiththis issue; theyremain
as secondary evidence strongly suggestiveof continued
gainbyindustrialcountriesinbargaining withagricultural ones; by
taking export prices Mr. Keynes does not really get substantially
nearer to measurement of the return to manufacturing labour.The
increased yieldofcorn per acre throughoutEuropeanditssettle-
mentsuptotheeveofwarisalsosecondaryevidenceonly: it shows thatwe
were still able to extend cultivationtonew fields and toincrease
theaverageyield.Itcould beargued, however,that this was only
happening atthecostof puttingmore labour intothe taskof
cultivation,eitherdirectly orin the form offixed capital-
machinery, railroads, andthelike.Butthisargumentisallbut impossible
toreconcile withthelast and most important fact in my
paper-ofasteadilyincreasing yieldper headofthetotalpopu-
lation,alikein Europe and in the countries settledfrom Europe up
tothe very eveofwar.ThereconciliationofMr.Keynes'views
withthisfactinvolvesaseries ofunproved assumptions so highly This
content downloaded from 164.11.203.58 on Fri, 31 Jul 2015 10:28:02
UTCAll use subject to JSTOR Terms and
ConditionsMR.KEYNES'EVIDENCEFOROVER-POPULATION I7
improbableinthemselvesasnottowarrantexamination.The total
population includes those engaged in providing the agricultural
machinery or fertilizers or the goods to be exchanged forthem. Mr.
Keynes, indeed, now takes the line thathe was speaking only of the
" industrial countries of Western Europe," and thatinclusion of
figures for other parts of Europe " vitiates" my argument.This
newpositionrequires examination.There is,asIpointedoutat Liverpool,
a theoretical difficulty, in these daysof nationalspeciali-
zationand co-operation, in considering thepopulationquestionfor
anyone countryor small group of countries apart from therestof the
world.Mr. Keynes' own method of inquiry-basedon the ratio of
exchange between industrial and agricultural products-isparti-
cularlyill-adaptedforyieldingconclusionsaboutoneorafew countries in
isolation from therest;his original inference from the unfavourable
movement which hethought he observed in his index
was,quitenaturally,oneofadiminishingresponse ofNatureto population
oftheworld asawhole,notincertain countries alone. Ifweare
considering particular countries,factssuch asIgavefor Britain,
atLiverpool, are far more tothepoint, and, as ithappens, are better
reasons for anxietyabout Britain than anythingthatMr. Keynes has
said;but as I also showed at Liverpool, the faltering of
Britishmaterialprogress recorded insomanyindicesduringthe Edwardian
agewasprobablyinpartapparentonlyandinpart real, but due to
transient causes.Mr. Keynes, however, still speaks of the
industrial countries of Western Europe, notofBritain alone. My
evidenceprovesthatthesecountrieswiththeothercountries on which
theymainly depend for food supplies, were between them producing
per head of the total population both more corn and much more
ofthebasicmaterialsofindustryinthesixyearsI908-I3 than in
i898-I903;this is true whether we limit our view to Europe or take
in the lands overseas occupied by men of European race and
typeoflife.Unlesswemaketheunprovedandreallyfantastic assumption
thatsomehow thewholeofthisgainwasabsorbed by landsotherthanthe "
industrialcountriesofWesternEurope," we cannot speak of a falling
or even a stationary response of Nature topopulation evenin
thesecountries bythemselves.Mr. Keynes' evidence
atanyrate-evaporatingasitdoes under test-doesnot help ustosuch an
assumption. The case, then, stands thus.There is no positive
evidence at all to support Mr. Keynes' views, whether as to Europe
as a whole or as to thewestern industrial communities
beforethewar.On theother hand,there isstrong, ifnotconclusive,
evidence negativingthem; so far as can be seen, material progressin
Europe continued to the eve of war at a hardly diminished rate.This
does not settlethe popula- tion problem for us now.It does no more
than clear the ground for a real discussion of theproblem.That
'prelimirnarydestructionof unproved assertions
isthesoleobjectofthispaper. Mr.Keyneshasdonegreatserviceinbringing
theproblem of This content downloaded from 164.11.203.58 on Fri, 31
Jul 2015 10:28:02 UTCAll use subject to JSTOR Terms and
ConditionsI8ECONOMICA population again toprominence.Butindoing so,
he has coloured withpremature gloomthewholeatmosphere inwhich
ithastobe discussed.Hehasseenthewriting onthewall,tellingthat" the
magnificent episode of the nineteenth century"was over, that even
before the war Europe's prosperityhad reached its climax, and, war
or orno war, wasonitsdownward slope.Hehasseenandhehas dazzled
theeyesofa generation.And after all,itseems, he was a
visionary-Cassandra,right or wrong, withno reasons for his fears.
Inour ante-chamber of war there was no writing visibletohuman
eyes;onthehuge highwayof material progress one nationmight gain or
lose on others, butall swept swiftlyon;there was nothing,
uptothemidnighteveof war nothing,tosuggestanyfaltering of
thegeneral pace,anysudden steepeningoftheroad. APPENDIX
NOTEONTABLEI
ThefiguresincolumnsI,2,and4ofTableIshowthecourseofprices,by
expressingthedeclaredvalueofthespecifiedexportsorimportsineachyear
asapercentageofwhatwouldhavebeentheirvalueatthepricesofthebase
year;thislatterfigureisobtainedbymultiplyingthequantitiesincachyear
bythepricesinthebaseyear.Fromi88Ito I9I3, I902(thelastyearin
ProfessorBowley'stables)istakenasthebaseyear;beforeI88Itheprices
ofi88ihavebeenusedtocalculateinthefirstinstancethevaluesforcom-
parisonwiththedeclaredvaluesandthepercentageshavebeenreduced
subsequentlytothebasis1902=IOO. Column3
givesthemeanofcolumnsiand2(textileandmetalprices),and
column5istheresultofdividingthepriceoffoodimports(column4)bythis
mean.Itshowsthepriceoffoodimportsinrelationtothepriceofexported
textileandmetalproducts;alternatively,asMr. Keynesputsit,itshowsthe
changefromyeartoyearinthevolumeofsuchexportsgivenforauniform
quantityoffoodimports.Toavoidcumulationofsmallerrors,Ihavetaken
thetwoexportpricesandtheirmeantooneplaceofdecimalsthoughIdo
notshowthesedecimalsinthetable.
Column6givesten-yearaveragesofthefiguresincolumn5, theaveragefor
eachdecadebeingsetdownoppositethelastyearofthedecade.
Thefirstcolumn, " Textile,"coverspracticallyalltextilemanufactures.
Fromi 900to1913thefiguresare basedonthosegivenbytheBoardofTrade
forCottonManufactures,WoollenandWorstedManufactures,andother
TextileManufactures(ImportandExportPrices,Cd.
Papers6314,6782,7432). FromI865toi
900theyrepresentanewcalculation.Broadly,everyitemin
GroupH-YarnsaiidTextileFabrics,ofthe " Articleswhollyormainly
manufactured"forwhichquantities aregiven
intheStatisticalAbstractsis
included(otherthanWool-FlocksandRagwool,Waste,Shoddy,Mungo,
NoilsTops;andSilk-Thrown, TwistorYarn),andistreatedseparately.
ThisgiveseighteenseparateitemsfromI865toi88ItwentyfromI882to
I889andthirty-sixfromi 890onwards,ascomparedwithfourteenin
ProfessorBowley'scalculation.
Thesecondcolumn,"Metals,"meansGroupAofthe" Articleswhollyor
mainlymanufactured"intheStatisticalAbstracts(" IronandSteeland
Manufacturesthereof'),withthreeitemsfromGroupB,namely,Brassof
AllSorts;Copper,Wrought orPartlyWrought-MixedorYellowMetal;
Copper,WroughtorPartlyWrought-ofothersorts.Aswithtextiles,the
figuresfromI900to1913arebasedontheBoardofTradeReturnofImport
andExportPrices,andfromi88itoI900on a newcalculation;in this, "Iron
andSteelandManufacturesthereof,"treatedbyProfessorBowleyasaunit,
issplitupintoseventeenseparateitems,makingwithbrassandcopper This
content downloaded from 164.11.203.58 on Fri, 31 Jul 2015 10:28:02
UTCAll use subject to JSTOR Terms and
ConditionsMR.KEYNES'EVIDENCEFOROVER-POPULATIONI9
twentyitemsaltogether.Beforei88ithesamegroundiscovered,butthe
classificationislessdetailed-withtwelveseparateitemsonly.
ThefifthcolumnrepresentstheBoardofTradefiguresfrom1900toI9I3,
andProfessorBowley'sfromi88itoI899.Thelatterincludedthirty-four
separateitems(assetoutintheStatisticalJournal,1897,p.438).Inmy
calculationfor theyearsbeforeI 88I,a fewitemshavehadtobecombinedand
afewofnoimportanceinthoseyearshavebeenomitted(Mutton;Preserved
Meat;Cocoa;Confectionery;CondensedMilk;Raisins;OliveOil;Palm
Oil).Theseomissions(representingabout2percent.ofthewhole)can
makenoappreciablechangeinthefinalfigures;twenty-fourseparateitems
remain.TheseriesfromI865to1913incolumn5ofmytablegivesasetof
foodpriceswhichhasatleastoneadvantageoverotherseries-ofallowing
forthechangingimportanceofdifferentarticles;theBoardofTradeprice
indexexaggeratesthepricesin
earlieryearsbygivingtosugarthesameweight
inI871,whenitwasdearandlittleused,aswhenitwascheapandcommona
generationlater.
TherevisionofProfessorBowley'sfiguresfromi88ito1902makeslittle
differencein thetextilegroup,buta
greatdifferenceinthemetalgroup.Thus theprice of
exportedmetalproductsinI 88 Iisgivenbyhimas go, whereas itshouldbe
I05, andthatinI89Iasiiiwhenitshouldbe99(1902being
treatedasIooineachcase).FromI900to19I3thetwosetsoffiguresrunas
inthetablebelow.Arepresentsthecorrectfigures,basedontheBoardof
TradeReturn,withitsdetailedclassification;BismycontinuationofPro-
fessorBowley'sEconomic journal figures.Thelastcolumncorrespondsto
thebrokenline (---) inthechart. MvleanofFood Import
Textiles.Metals.Textiles anidPrice asper- Metals.centage of
A.B.A.B.B. Columnn (5). (l)(2)(3)(4)(5)(6) I900...103 102II8III 107
93 I9OI...103103104 Io8Io695 1902...100100100100100100 1903...101
103 101 103103 97 1904...IIOlo895 103 Io693 1905... *I
II1109510310794 1906...II8II8100 104 III91 1907...125125I08IIIII890
I908...120122101110II693 I909...I I5 I I697lO9111299
19IO...1241269411211994 1911...129130 95I i612390
1912...12712910012112594 1913...132134io613213387
From1902to1913themeanofthetextileandmetalprices(basedonthe
BoardofTradeReturns),andthepriceofwhollyormainlymanufactured
exportsas a whole(drawn from
thesamereturn),followmuchthesamecourse.
IthasnotseemedworthwhiletocarrythelatterseriesbackwardsbeforeI900.
Atestcalculationcoveringallmanufacturesotherthantextileandmetal
forwhichquantitiesareavailableinI88i,yieldsinthatyearapriceof I05,
(onthebasisi902ioo).Thisisthesameasforthemetalproducts.The
exportsincludedareAlkali,Gunpowder,Bags,Bleacching
Materials,Candles,
Cement,CordageandTwine,Glass-Flint,GlassBottles,Glass-other,
Leather-TannedUnwrouglht,Leather-BootsandShoes,Zinc,Oiland
FloorCloth,Paper-Writing,etc.(excludinghangings),Paper-Unenumer-
ated,SkinsandFurs-BritishForeignDressed,Soap.Theseexportsini88i
This content downloaded from 164.11.203.58 on Fri, 31 Jul 2015
10:28:02 UTCAll use subject to JSTOR Terms and
Conditions20ECONOMICA
hadadeclaredvalueof/12,253,000,lessthanone-tenthoftheexported
textilesandmetalproducts. NOTEONTABLEII
ForGermanythenumbersfromi88ItoI91IshowninTableIIhavebeen
constructedfromdataintheAbstractsofForeignStatisticsonthesameprin-
ciplesastheBritishfiguresinTableI.Thearticlesincludedareasfollows,
withthedeclaredvaluesin1902:
Textiles.-CottonYarn;CottonManufacturesSilkManufactures;
WoollenYarn;WoollenManufactures./38,800,ooo.
Metals.-Iron,Pig;IronBars,Malleable;IronWares,Coarse;IronWire
(fromI884)./14,600,000. Other
Manufactures.-MachineryofallkindsincludingLocomotives;Musi-
calInstruments;GlassandGlassManufactures;Paper;Leather-Dressed,
Dyed;LeatherWares,exceptGloves;Skins-Dressed;andfromI884,in
addition,Dyes;WearingApparel;andBooks,Maps,Engravings,Litho-
graphs,etc.,47,000,0 00.
Food,DrinkandTobacco'Imports.-Wheat,Barley,Maize,Rye,
Oats(groupedas "Grain,FlourandMeal"beforeI894);Coffee;Herrings;
Tobacco;Wine. ?49,400,000?
GermanFoodImportsdonotrepresentnearlysowidearangeofarticles
asBritish(excludingamongotherthingsCheese,Eggs,Bacon,Butter,Tea,
SugarandMeat).Graindominatesthemaltogether;unfortunatelyforgrain
beforei 894 thefigures for
differentcrops-wheat,rye,barley,maize,oats-are
notshownseparately;apparentchangesofpricemayrepresentonlychanges
intheproportionsofthedifferentcrops.Toogreatreliancecannotbe placed
onthesefigures.
Incolumn(7)accordinglythepricesofGermanmanufacturedexportshas
beenshowninrelationtothepricesofBritishfoodexports;thatis,itdiffers
fromcolumn 5 inTableI,onlybythesubstitutionforcolumn3 ofTableI,of
column (4) fromTableII. Themovementof theMean of
TextileandMetalPricesafter1905is remark-
able.Uptothatyearithasfollowedthegeneralcoursetypicaloftheother
indices,bothBritishandGerman,fallingtoaminimumaboutthemiddle
'nineties,risingtoa markedpeakinI900and
fallingagainwiththetradecycle.
ButwhereastheBritishindicesbothofexportedmanufacturesandoffood
importsandtheGermanindexof" other"manufacturedexportsthensoar
upwardsto19I3,theGermantextileandmetalindexfallsviolently.Thisis
averysuspiciouscircumstance;probablyitrepresentsachangeinthecom-
positionof theGermanExportsor of theirclassification.Itis an
addedreason fortrustingexportandimportpricessparingly. This content
downloaded from 164.11.203.58 on Fri, 31 Jul 2015 10:28:02 UTCAll
use subject to JSTOR Terms and Conditions