of approximately
1 .5
millions output.The questi on arises immediately; if there is such an advantage in terms of c ost , by curtailing marginal suppli ers , why is this not practised more? The answer appears to lie in the fact, that this curtailment is happening - not deliberately in most cases ,
(
although many factories insiston frontier suppl iers bringing their cream to a suitable loading point
)
-it is happening largely in the process of time . Supplier density is increasing all the time , and this has the indirect e ffect of offsetting the high haulage rates which arise from collecting relative� small offerings from marginal suppliers .
The average cost of hauling cream is greater than that of t otal wages , but in the course of time , wages tend to increase and that of transport tends to decrease largely due to greater supplier intensity which offsets the e ffect of distant marginal suppliers.
4.8
It is this process, together wi th thepractice of contracting out the cartage of marginal cream, that has enabled transport costs to decline through time.
In
every instance where managers have declined to operate beyond a certain distance from the factory(
althoughaccepting cream at the supplier ' s expense
)
, cream haulage costs have beenkept at a lower level than unit wage costs , thus indicating economic efficiency has operated in that haulage costs have been substituted for other factor costs .
An analysis of the relationship between wages and haulage costs c onfirrre this. Some
23
factories(
of90
examined)
had cream haulage unit costs under70
per cent of those of wege costs , while23
factories had haulage costs inexcess of
1 30
per cent of unit wage costs. Of the first23 (
low haulage cost grrup)
,1 3
factories lifted cream from within a1 5
mile radius and only4
exceeded a30
mile radiuso Of the23
high-cost group,1 5
factories exceeded30
miles , and8
exceeded1 5 miles from the factory.
Of the smaller fact o ries(up to
2
million lbs. output)
, 44 per cent had no supplier more than 1 5 miles from the fac tory and 44 per cent had costs under 70 per cent of wage co sts.48
The decline is relative to total costs . In one factory where condit ions were relative� stable , and output over20
years varied between5
to 6 million lbso per year, wages increased :f'rom1 0
to1 7
per cent, while Cream Haulage Total Costs decreased from20
to 1 6 per cent for the same·period.(1 931 -1 950) .
Of th� larger factories (exceeding 6 million lbs. output) , 67 per cent had suppliers in excess of 30 miles from the factory, all of which had haulage costs higher than wages costs ( of which
37)�
exc eeded130;6
of the wage costs ) .Generally spe aking, where one-half of the cream was lifted within
7.5
miles from the fR.ctory, cream costs were lower than unit wage costs, but where one-half of the cream exceeded7o5
miles , unit haulage costs exceededwnge co sts proportionately as the
50
per-cent-point increased in distancefrom the factory. 49
Other factors The main elements in cream
haulage co 3ts are weight, mileage and tine. From a theoretical point of view, weight asswnes a peculiar function in that it is not diss imilar to a kind of
" fixed" element in cost, whereas mileage and time ca...'1 be compared to "variable" ·elements. While haulage costs would be meaningless without weight, from a
point of viP.,., addi tion.3� weight adds very little to total costs over
a given
fixed
route. Tha t is , a large part of haulage costs will be involvedeven if no cream were lifted at all , for the fixed expenses would be the same
A1ld the variable expenses would be slightly less only.
Some 26 factories which used their �vn vehicles exclusively were examined for cost of transport data . These factories were carefully selected as repre
sentative of' the entire range of output and terrain over which collection may be made. The relevant datA is shrnvn in the follm� table : -
49
�ee the data given in Table XXX (p. 1 44) , where some of the highest (best) petrol c a1sumption data occurred when the loads were theheaviest , and worst when the loads were lightest. Reading conditions affected petrol consumption more than loads.
TABLE XXXII. D. __ . . .. ,. . .. �
RELATIONSHIP BET\{i';EN .. TD ·E1 .EIGHT !UID CO:JT OF DA'.r A
--
OF FACTORIES 'rHE SK� :JON
.
ofNo�
.P..:;rer;·,.··e
. . ;;ir·lltere·
•jj} lifted____ ... . �E>. .J----· - ---'-- ""· Ne:i.. ht
.'cV"era
·re Avo An.n& '\.vG UP.it Unitf'�:,ctor- of ere :l!H lifted ··nnu·�·� · cost
_£?._; ;t
T 'OOOs
---('OOOs
!b2:
ies
lb.s . )
___ ...._ mile::: .-read-
Cost.<:> £ .. £. d.
.££_
d.c(1 )
(2)
(3 )
(4)
(5 )
(6)
(7)
(8)
(9)
Up to2,000
5131 1
24, 934
52
636
1 41 6 o.6o1
0 . 1 97
2001
-4,000
4
3235
49,406
68
1 lf-1 83281
O o 650
Oc5
27
4001
-6,000
44390
66 , 21 0
67
2260
4455
0. 606
o. 66o
6001
-9,000
4
821 4
97
, B21�o 84. )2816
374 O .. l,.76 0. 976 9001- 1 2,000
3
(9273
1 1 2 ,4.8783
3701
7204 0.462
1 .071
1 2001 Qv . . L� 1 8,0003
1 l;..:l
529
1 50,767
97
5
1471 1 , 1 5 8
0.408
1 .479
1 8�000
3
23 ,
089
1 66, 1 67
1 39
67C,8
1 6 ,361
0
.. 426 2o71 2 Column(
1)
"hems cl.'l s:·.:-interv8ls of crr-·::•m lifted to the nearest 1 000 lbs . ColUIJin(3)
sho1N:'.l the wt::ighted l':!Ver:·,�e of Column (1)
., Thus in the first class-interval five :f'::Jctories lifting up to two million lbs o of crr·.r.m <trr ·recorded,.
The cl� s s-interv· :l�' of Colw:m ( 1 ) progr·" � ; s in heterogeneou�1 interv:-:..ls in order to bring out the p;;.ttern of rel atiom:hips morr::· clo::;P.ly$ 'rhu.'<· ·re:r-E:9 or
1 0
fnctorie s in the: l r1 8 t c b ss-int•':rY<-�.1 !ihich might h:·vc been include:dbut w er·e deleted beca.l\::J e
o
f doub t "lbout cert::Jin de::tails . ;;'hen ex&mined .sep<::r<·tely,
hmrever, their averP.ce figures c".pproxima.ted very closely those of tbe.:tbove t�ble in .:;very resp,�ct .,
Column
(�.)
shov·1s the weighted :t'i!er:;.ge .mnua.l mileEJgr-: run on cre aw collection. Column(5 )
is � co-efficient or ratio computed as Colu111n Colwnn t4) ..()J.
Thi.·. r(" tio i::; .-
,
·n•::a. ll''r. vf th( m:illber of lbs . of cre• •m lift "":d per ·lil-:;; and is an indicator of the inde}?endent V8.ria.ble, weight., The greater this co-aef'ficient , the lov.cr shoulCJ. be the unit cos t(oth,_r
th:L."lgs beinsequ.:,�).
Coltunn
(6)
shov>s the weighted average "'nnua.l .�age-bill 8 lloca.tE:d to creto.m hauh,ge and is �) relative l11.".: 8.sure of Time consumed on cre2.m collectionoC olur.nn
(7) shCI'NS
the ''veighted average of the am:ua.l cre:::>rn collection co.sts s It includes the Clatn of Colwm;(6)
and tot:::·l truck h2.ulnge co sts . Column(8)
i�:; the weighted average of unit cream collectior' costs in ter.rns of pence per lboB/F.
Colu:�m
(9)
is a derh•.s tive of Colunm(i'l)
:md Colw rn(3 ) ..
It assumese. constant average weight of cn:-am lifted of
4,000,000
lbs., per znnum. Inthe first cle.S (">"""intcrval in C olWilll.
(.3)
1 1 ,31 1 , 000 lbs . of cret:·m were lifted. been lifted, thenob,;i.cusly·
the [�verage Ibd the hypothetical 4-, ooo, ooo lbs.unit cost vtould have been '�.ooo of 0 , 601 , i. e. 0., 1 97d,. ""' �o CoJ.t;lr.n
(9 )
unm:::..sk!: the real efficiency of cream c ol lec tion of the larger fActories . It putsall f:,1ctories on the S8ffie basis B S ree;::;rds wei��ht of cref.'m
-
in other words 1i t holds the V8 ria.ble element of "weight11 (:onstant, and this tends to reduce unit c o1:te for· low-ont:rut f;:;ctories , ,"1•1rl inoreo se u..rrlt costs for high-output
factorie s
(as
revealed in. Column(5) )
. It should be observed here , thatthere wa.s consider1:.ble dispersion cf all indiv·i
d
ual observations from theme:n n of the class-intel"V::Jls,. This dispe:c2ion WP. S quite mt:.rked
when
the ir"d1.- 1 o.L' s .1. .. 1. (r.:, ') -.� o -:: (;e � .. d 2 eo l t l 1mn(0)
.,� , a nd th ose o(4)
agrdnst Column(8)
and Column(7)
�:tgainst Columns(8)
o�c·(9) .
In fac t ,th e only sc.,.,tter d.i�Er1'l:rn v;hich sJ1owed 2n;y· <<.p:parent rel at i onship, w:� :o thn t
obt:d.ned froH: plottin£ tl::·e indiviCiur:,ls of Colurrn.s
(4)
::>nd(7) .
Here 'l ,;-.;:·rkecltre: 1Cl c:f clc :'Ocly clt:.ste:ring dot[: w::; s observed up to abcut 1 01000 , 000 lbe .
of
cream, and this contirrc1ea in a more di s.persed Vi8Y with incre,<J seo poundage :.md ccsts. It denoted a marked relationship between rrdlez.ge and t otal costs .By aY6r<J.gin.g
the::oe individuols , other r·elation:;hi:p:: c :_;.n be ob�:er1Nd . �o:f
·rhis
must .s.sswnt.'; other things held equal-
the s Gme dist�mce covered ,and the takf;l1 tjJi;8., GiY· \': G fleet of truck:� , cream h<?.Ul:::ge
cc 2.tc· e.r� Q.erivative of rd.le2g� , time ;md vtc·ight. l•'!ilease nnC. time
8re the b·.ro v.:•riables nffecting the d ispersion of cosh� nv.:-stlyG
We:i.ght tends to be a const<:!nt.. It
tz.ke;!; just as
W<�rv i:-J.les ��nd timeto coll e c t
1 00
c.9 cont::J ining 1 0 l'bs . of creama.s
1 CO cans containin&::; 20 or
50 lbs
.. of cream(
a�;suming the cans to be in identicalplC�ce:=:
etc.) .
But
obviously the30
lbs. cons will shmv unit co st1::just
one-thin:l of those '"·ith 1 0 l'o::;c �.,�eh .. 'Unit co:c: b:: cBmouflageefficiency. If TM.ny large f<-lctcries suddenly had their roun<lage cc.rt by one h;,lf, tL�? ir unit ,;o�t�; v,'ould aouble. Yet there :::;re; f"'ctories
':·•i th .3 r, output of approxiJr�?,tely 500 tons covering er. collection are2.
identi.c2.l in si z�� , nrea r:md di sp."rsion �l ;': :;ome f.<wtories 1dth 5· 1000 tons output, <·.t le.1� tbrcn r3.ouble tb; u:ut �:o:o: to C ol umn 9 iE l1s eful
<� s i nd icating th<· ·ua.gni tud '' of thi c han � P. , but p ract i c o.lly tr:nd :J tc
There is a steady increase in the density of cream lifted
(
see Column(5 ) ) .
This should not be interpreted as meaning that small factories pickup about one-half the quantity of cream per mile ,
(
for example)
of those in the second last class-interval . This would be so if we could assume constant density per mile for each mile out from the factor,y. Quite a number of the larger factories picked up well over1 , 000,000
lbs . of butterfat within the first five mile radius . A few picked up well over three million pounds of fat. The smaller factories picked up in the vicinity of200 1000
to300,000
lbs. of fat, while medi�sized factories were often pressed to reach the half-million lbs . r.oark. That is, the fat a.ensity ratio is about 1 : 8 between the SID2cllest and largest factories for the first few miles . However, larger factories extend out relatively much further than small fact ories and tend to go con.Siderable distances to pick up the last 1 0 to 20 per cent of their cream, otherwise their weight-distance co-efficient would be even larger.The relati onship between
Columns
(6)
and(7)
is fairly constant, but there is an int eresting relationship between Columns(4)
and(6) .
Column(4)
measures time indirectly,(
in hours)
for if mileage given is divided by1 5 ,
we ha.ve an approximate measure o f tiloo taken away from the factory on cream colle.otion. Colwnn(6)
measures time paid for, whether consumed on cream haul age or not. For example , in the first class-interval,24, 934
miles at1 5
m.p.h.indicates approximately 1 662
hours were taken on cream collection, which, when divided. into the aggregate average Wages, yields roughly 7/6
per hour. Taking penal rates of pay into consideration, thisme
ans one-tr.drd of time of driversis not on cream collection. However, as scale increases , it is clear that
these hourly rates tend to increase s o that in the sec ond last interval
(
for .51
example
)
, approximately1 0, 000
hours receive roughly£,5 ,000
o r1 0/-
per hour.51'·
·This is based. on the fact that the average driver with penal rates earned approximately
£520
per annum or57-
per hour. Depreciationof drivers ' accommodation was charged in with wages occasional�,
but apart