Class I Differentials with $2.50 Fuel to $5.00 Fuel*

Chuck Nicholson & Mark Stephenson Cal Poly San Luis Obispo & UW-Madison

Living on a Roller Coaster: 18'h Annual Workshop for Dairy Economists and Policy Analysts

Outline (Surprise!)

• Class I "Price Surface" with different fuel costs -Work funded by AMS Dairy Division

-The main topic

• *Costs of "Localizing" the US Dairy Industry - Recent Food Policy Paper

• *Use of USDA and CBO Price Projections -Recent survey for Ag Policy course

* Special bonus sections! Or " 30% more presentation FREE"

* EXHIBIT

I 3o I

1

Outline

• Background • US Dairy Sector Simulator (USDSS) basic

description

• Modifications from original USDSS model -Additional final and intermediate products

-Nonlinear yield functions

-Revised plant locations

• Class I price surface results

Objectives

• Update the "cost-minimizing" price surface to 2006 -(Okay, so now we should probably update the

update)

• Examine the impacts of changing fuel costs on the Class I price surface -From the 2006 diesel cost of $2.36 to $5.00

-At the start $5.00/gal diesel looked big!

-May 2nd: $4.12/gal, with CA at $4.47/gal

2

Methods

• Cost-minimizing model of the US dairy sector

• Long history of this analysis -Pratt et al. 1998 Class I differentials study

-US Dairy Sector Simulator (USDSS)

-Also funded by AMS

-Input into the differential structure under order reform

• Model was given an "extreme makeover" in 2008

USDSS Model

• Spatially-disaggregated 'transshipment' model

• Minimizes the costs of milk assembly, interplant (intermediate product) use, and final product distribution -Some versions have included processing

costs also

• Nonlinear programming solution technique

3

USDSS Model

5 product categories

US Milk Supply Locations Given milk supply locations ...

f• .I

i ) • • •

~ . \.· , ' I

·.

4

US Cheese Plant Locations ... and processing plant locations, like cheese ...

• , ' ....

...

.. .. .. .... .... 4o.

....

l ,..

. .. .. ..

.. .. ..

......

..

" ~·· ... ·~ .. .. ~ ... ... .............. ... .. .. :;~ .... ... ... .. . .. .. .. :\ ... .. .. ...... ._ ... • •r;. • .. ~ .. .. . t i "' .. ....... ...,. ... .......... ··.~ .a. A A A "•i J

... ... ... ~ .. .. . ··"'" ... ..

-.... · ....

...... .. \• ' '

US Cheese Demand Locations ... and demand locations ...

t~ .--~! ~ ··-- - I , t\ I • -· \ • . • • ., ',

I . / l.. • • • ':..• - :;-i-1 •• ~ "t. \ t • r • ·' • •••• ·r~ f .· ·'

I --L... .:_1

' . • I. ·· ·· •. 1.1 .~····tc~ . ' • ~~ , - 1 • 0 o o rf •' oJo ; •.- •

\. ; \ II '. .: J • ·.-. :\ •• :: : , ·:-;;:· ···.1· .. · ···:·r. \ ~. I • · ) :: •l'; • 1 0 J ' -' ~t:; • •, 1' • : I • . ) • • : • J • { ." • \ ~ , •r' - f._. I • , , \(t/'• , • ', v,

- • ---r:::-1 . • -\:. \ :·: ;/. ·.· :=·~ :.. l • I . I . ..._. . . . . ~~ I' -~ : . • • . · ..

', . . ·I • . ... " .... : -·. - . . .... \ . ~ . . . . :

. . . I I . r \ ;. ··~·· ~L· ... ~··.\ "' . . , . ~ ' . . . ~) ~;;.·. . . ~ l ..

5

The USDSS Model

• Figure out least overall cost for ... - Farm to plant (milk assembly)

- Processing at plant (products)

-Distribution to demand

~~ . -_ ..... $ $

$

Model Solutions Represent: • The work of an imaginary "Dairy Dictator"

-Solution is least-cost for entire system -No emotion about keeping plants in

business

' . .

- . -·~-, ---·· ~~"~

.. ;m ~ ' · ~·:- ~·''\\~ ~ ,r,;-,~1 ·'\.. • ' \ •

""~ '·--..'!

&1 I ·", .'

6

Model Solutions Provide: • Insights about some incentives for a

plant to be at a location -Not actual outcomes -Many important factors omitted

• Uses data for two months -May and October -Original study used 1995 data -Most recent update 2006

Model Solutions Include: • Plant locations • Milk assembly flows • Product distribution flows • Spatial milk values

7

Revisions to Previous Model

Limitations of the previous USDSS: • Limited product/component disaggregation

- 5 products, 2 components (Fat, SNF)

• Newer products not represented -Dry Whey, WPCs, UF milks, MPC

• Fixed product composition (proportions) • Least-cost solution

-No 'policy'; shipment patterns not consistent with incentives for pooling at all points

Changes from Original Model

• Expanded number of final and intermediate (interplant) products

• Nonlinear yield functions based on use of cream and skim fractions and intermediate products -Results in endogenous composition

• Modifications to plant locations -Reduced fluid plants, added key locations not

. . In preVIOUS

----- ······---------------

8

Revised Model Final Products

• Fluid • Dry Whey

• Yogurt • WPC34 & WPCSO

• Ice Cream • Lactose

• Cottage Cheese • Casein & Caseinates

• NDM • 4 MPC products

• Butter • Dried buttermilk

• Cheddar Cheese • Other ECD

• Other Cheese

Model "Intermediate" Products

• Cream

• Skim

• NDM

• Ice Cream Mix

• Fluid buttermilk

• Dry buttermilk

• Fluid whey

• Separated whey

• Whey cream

• Dry Whey

• WPC34 & WPCSO

• Casein & Caseinates

• 4 MPC Products

• Condensed Skim

• 4 UF Skim Products -Permeates

Intermediate means used in the manufacture of another dairy product

9

Model "Intermediate" Products

Example of IP flow

NOM can be used in:

• Fluid (CA)

• Yogurt

• Cheeses

• Ice Cream Mix

Another Example

Separated Whey uses:

• Dry Whey

• WPC34

• WPCBO

Fuel Cost Scenarios

• Changes in Class I price surface with changes in fuel costs

• Three fuel cost scenarios: - $2.36/gallon diesel

- $3.50/gallon diesel

- $5.00/gallon diesel

• These average US diesel prices were modified to represent regional cost differences at origin points for shipments

10

Results

May 2006, $2.36

11

May 2006, $3.50

May 2006, $5.00

12

U.S. Coui'IUOI (GonoJ••II•tOd)

MII)'2J6 - 1100>;10Co•'llt't.GOO

- U0l0t UC«-30 .,DO:I)O, UU.'OOO

]1((11001· ~ '~0

U COJ01 · ~ltQ))D

- ~00001 &ot'('Olo

-e.tCb:lO!&.KC\).)0

- UO'.m! eHC(COO

U.S. Cou .. Uol (Gonorollzod)

MII)'J50 - IICXJI000•"(7!lCOQQ

- 2ll)))Ol'"""'' - l.KOOCI\ UOCOOO

llt0001. ,f,. J OC()OQ

HIX»!H • $ 100000

- ~JCOOOI OOOWJIJ -tWMtrtt<IOOO

- ''rooo' • •t~:ooo

MAY 2006, $2.36 Fuel Cost

MAY 2006, $3.50 Fuel Cost

13

U.S. Cou•tloa (Gonar.~ll:ud)

llly500

- '1o.»GC·1'1«0ll

- 2.l(l)l01 >t«"ll -:21CDX11~...CCOI

,ICX))O, ·• •ocoo• l &O:n; 1 • S 101m I

-i-!0:1011al!.ll..cot

-e,u,o::.o, r1oc®e

- IIO'ne1 & 4~f

MAY 2006, $5.00 Fuel Cost

May 2006, $2.36 to $5.00 Difference

\)

I

14

October 2006, $2.36

October 2006, $3.50

15

- tXOIJOO . li'CIIlOOO

- tlmllh >9«001 - 2.100001 ).GW)ljt

]KlOQ01 .... .&(((01

~ 4C:OJCU • UOOODt

- ucuocn a.xax~t -t..t:<I.I001rtCI«<It

- '100001 a •ocoo•

October 2006, $5.00

OCTOBER 2006, $2.36 Fuel Cost

16

. ,,CI.)D00.,1CGOOo - 2.2COXI1 tlCCQlO -l.to:loollO«<DO

J liCOOCI'· " ~cnooo

.:. 'C0001 · s 1t(ffl0

- S.201)o)QI SOOOXht

-U:Cl.JI>alr!GCQJO

-/IDMQI lll'loCOOII

Clu e I

- IJOI)OOO.lXIIXOO

- ::!,1»)(11 ?9tt«<fl - l .IOI.;J001lob...'O:Xlll

l Kl0001· "· 'C0:00

4400QCI1 · $o100000

- UC\1001 O.o«.o::O

- • 000001 r 100000

- I IC0001 .;1((()00

OCTOBER 2006, $3.50 Fuel Cost

OCTOBER 2006, $5.00 Fuel Cost

17

Oct 2006, $2.36 to $5.00 Difference

Key Results

18

Can Also Examine Flows

• Movements of products

• Fluid plant inflows and outflows

• Cheese plant inflows and outflows

19

May 2006 $5.00 Flows, Fluid

May 2006 $2.36 Flows, Cheese

t ~ -

20

May 2006 $5.00 Flows, Cheese

Key Results, Flows

21

Now, Localization

f ood Policy 000 .... ,

The costs or increased localization for a multiple-product food supply chain: Dairy in the Unirc:d States Charles F. Nlchol!on u. ll!l,url l G6m(Z ~·. 0 liV(f H. G.lo"' ·n.--;-.. oofio-•••ulfur.-1'4oowl'wo, .. ,.,, itw\11•""'"' 16-lL.utM:..,..c;.o~UW".,_,."'-' "L-..hfl ~~'N(''\"flrlt-\W\._.,.._{-.-t,.af"!h "-' " '' ..... l "?'II""" ~--1<':..-.~1-••--.t• .... -- .-.u.. .......... ,.t .,.,1 .. ' 1 _. .......

........ _...,. ~l:~!CII ~.~ : ...... .wJ ko.:,& o&.. .. 9'f'Vrll~1o) .-.,-... ..... -.~ ............ \.o~ ....... -.......u .,.._ v.:••.or o ............... , ,,_ ... .. ....., .... --~.

n., ·-.,.-... ·~ ._,, ..... ~r...., ... ~.(Oof' •lll ~ t,., ... ')oiJ.otlVf\101'Nt' ~ ..... ,.,. ,...,LI ol......,.,~ _.,;n t'll:l tiQPf ~ Wlh. AnnW'C ~ d.'a h., ,..po. t:tlf'llllt:'* fi'C'd.wl WYhl'""ftltOjiiiiiC ...... t\OI~..,IJwb-.l'ldlftaln •• t.'wt #V:y.-~II'MIIIp.a~ll_~ N tf.fll• ... ~t ~koiNU"'C,..f')'o!'\VII~' 11!"-'"'C"ftclil.al~ ..... ,.ttl .. llllllo.oio ............

li)" ~nUlL~ U.r.tp:na~Md ti:\:l!Wllftn~UIIft W~r.~t;JIIf;lf ~<CJ: nWwfV•I _,_.., w.,...., 10 ,,.,.,..,, 1t!t'IT\oKiw "C'N'~ c: ~••C'..,., ~ .. ~ .. , ... tuu ... "• rn.., ..... ,.r.'l)illl.ft ~ilnlltltD illuu.Nl!K.&.kr.U...., l'WM~ft11~U-n~r•IN;IP\'U'.U.

Olll t.l*'(l Wg ... ,~·.,.~- .. ,,,,."r.'"'.-,.."' ""'"C...,..U l'f"'IIOIt\.I!Nifll-'•-"" wt•"lllc..tt lillf".;wd l1udo1~ ""'OIItn .......... ~ ~U Wllllor '"'"'ol>C ~"olrC .u:d dhUIWII;t~~~ ~'- (W ll".tft'IU•Mt ll/1( f lllr(l;o .. IWIIl \ll"~tlrTUw,:i!'~ •JIRI!i("t.IMI!ilf~ lnrlt'i'('tlli'-"'('rtC't' I•" It ' • ,,.,-.,~..,. ..... ,. nln.r~r~•~"".,..,w~..r~d..My ... .,~IO.u•~· .. "Mtto..Jh .... ..,.,~..c-.h .. lftm1 d CMI Ulli! 4..uuno II'""'"' fhr NiAOt•!'lii;:J llortW'""' lf!UU...J ko.:.Jau'- U'1lil n-th l .... '""

IOI:ll:"illtt'1f.W,f

Food Industry Localization

• Given lots of positive talk -Reduction in food miles is often a goal

• Relatively limited analysis of the costs or the benefits

• Usually described on at a "local" scale - "Money stays in local communities, helps our

farmers"

• (Not really consistent with trade theory)

22

Objectives

• Examine the costs of reducing "weighted average source distance" for the US dairy industry -For all products

-For fluid milk (highly visible consumer product)

• Examine the limits too WASD reduction with current spatial organization of the industry

• A "systems" approach, not a local one

Methods

• Use the USDSS!

• Scenario 1: Add a constraint on the total WASD for all products

• Scenario 2: Add a constraint on the total WASD for only fluid milk

• Minimize costs given these constraints

23

Scenario 1 Results

• Maximum feasible reduction is about 20% -64 miles (from base of 317 miles for all)

• 20% decrease in WASD increases system costs -2% (May) to 4% (October)

• Assembly costs are decreased (up to 14%)

• Distribution costs are increased (up to 24%)

• Redistribution of costs in supply chain

Scenario 1 Results, Continued

• Optimal plant numbers are lower

• Some products decrease WASD -Fluid milk butter, yogurt

• Some products increase WASD -Cheese, NOM

24

Costs of WASD Reductions

20

15

Q) 10 Cl 1: .,

5 .c: (.)

~ 0 0

-5

-10

100

50

E 6 0 IJ)

~ > -50 1:

~ -100 1: ., B -15o

-200

- -lnterplants Processing Distribution Total Costs

Cost Category

10% WASD Decrease 20% WASD Decrease

Product WASD Changes

Fluid milk

10% WASD Decrease

American cheese

Product

20% WASD Decrease

_I Nonfat dry

milk

25

Scenario 2 Results

• Maximum feasible reduction is 12 miles -From 112 to 100 miles (11%)

• This reduction would increase system costs by 12% -Lower assembly but higher processing and

distribution costs

• Average US gallon cost increase $1.66

• Would increase total WASD by 30% - 98 miles, to 415 miles from 317

Summary

• Reducing WASD in dairy has limits with current spatial organization

• Total cost increases not large, but re­distribute costs in supply chain

• Overall WASD reductions do not reduce for all individual products

• Costs for reducing fluid milk WASD are higher

26

Finally, Price Projections

• Cal Poly Ag Policy Student:

• "Price volatility won't be a problem in the future. Just look at the USDA and CBO price forecasts."

US All-Milk Price Projections

20 19 1 ___ _

18 17 -

i 16 ~ 15 .,.. 14

13 12 11

10 r '

2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

- USDA -CBO

27

Finally, Price Projections

• Cal Poly Ag Policy Professor:

• "Don't believe everything you read without understanding where it comes from and what it is used for."

Projections: Trusted? Used?

• 4-question survey

• Sent to workshop mailing list

• N = 79 responses (Thank you!)

28

40

35

30

25

20

15

10

5

0

How Well Predict Future Price Patterns?

Not well at all Somewhat well

•USDA CBO

Statistically significantly different (despite similarities in forecast)

-Quite well

Aggregated percentage > " Somewhat well" is < 10%

50

45

40 35

30 25 20

15

10

5

0

How Much Use for Decision Making?

Not at all

Statistically significantly different (despite similarities in forecast)

-Somewhat Quite a bit

•USDA CBO

Aggregated percentage > " Somewhat" is < 5%

29