CHAPTER 28

MANAGEMENT ACCOUNTING SYSTEM DESIGN

Changes from the Twelfth Edition
All changes to chapter 28 were minor.
Approach
A brief summary chapter seems to help students consolidate their previous topic-by-topic learning, and
they appreciate such a chapter for final exam study preparation. The summary of the many different
adjectives placed before the word cost! and the concepts behind these adjectives is useful. "owever,
because there is no conceptually new material here, one need not spend much time on the text. #ather,
instructors can jump right into a discussion of one of the chapter$s cases, which is intended to raise issues
that cut across multiple chapters.
Cases
Puente Hills Toyota raises many financial responsibility center, performance measurement, and incentive
systems issues in a car dealership setting.
Axeon N.V. illustrates the real world application of many management accounting and control concepts,
including incremental cost analysis, capital budgeting, sensitivity analysis, transfer pricing, organi%ation
design, and control system administration.
Case 27-1: Puente Hills Toyota
Note: This case is unchanged from the Twelfth Edition.
Purpose of Case
This case can be used to motivate discussions of a number of topics, including financial responsibility
centers, performance measurement, transfer pricing, and incentives. The setting is an automobile
dealership, a business about which all students have some interest and understanding. And the setting is
real, so students can benefit from secondary learning about the industry and business.
Suggested Assignment Question
&valuate the performance measurement and incentive systems use at 'uente "ills Toyota. (hat
changes would you recommend, if any)
Discussion
The case illustrates clearly that financial responsibility centers exist in great variety. (hile textboo*s
describe the possible generic responsibility centers as cost centers, revenue centers, profit centers, and
investment centers, this case shows some of the variance that can exist within these generic categories.
The general sales manager is held accountable for net profit. +bviously, the dealership also *eeps trac* of
balance sheet items, but the dealership general manager$s incentives do not seem to consider them. ,o is
the general manager an investment center manager)
#egarding performance measurement- The departments are profit centers, but not all costs are allocated
to them. They are more li*e gross profit centers.! The salespeople are held accountable for gross profit
on the deals they initiate, so each salesperson is also a little profit center. The service advisors are paid on
commission, so each advisor is a revenue center. The service technicians, though, are paid for wor*
accomplished.
.t is useful to discuss why some seemingly uncontrollable indirect costs are allocated to departments /see
&xhibit 01. These allocations are mandated by Toyota, so that they can compare dealership departments
on a common basis that treats each department more or less as a standalone business. Allocating the costs
also gives the department managers information as to what services are being provided for them, and it
gives them some power to complain if the si%e of the allocations becomes too high.
This issue can lead into a discussion of the differences between authority and responsibility
/accountability1. 2enerally, managers are held accountable for results in areas where they have authority.
3ut here, the general sales manager$s bonuses are based on performance measured in terms of profit after
overhead allocations /line 45 in &xhibit 01. This is an example of a situation where this manager is held
accountable for areas over which he has less than full authority. 6onversely, the service manager$s
bonuses are based on the department$s gross profit performance, which is before allocations.
.t is apparent in the case that some of the measures can be gamed. 'uente "ills Toyota managers worry
about the gaming in the service area, and they seem to have ade7uate controls over these behaviors. The
6ustomer ,atisfaction .ndex /6,.1 measure is also gamed. 3ut here the managers seem to encourage the
gaming because it ma*es the dealership loo* better to Toyota. 8oes the dealership need an un-gamed 6,.
measure for its own management purposes) "ow will they get accurate information about problems if and
when they exist)
Transfer pricing of service jobs done for internal customers /particularly the used car department1 is
done at mar*et prices. The alternative would be to give the internal customer a discount, or perhaps even
to transfer at cost. (hat would happen under that alternative) .t would shift profits from ,ervice to 9sed
6ars. .t would also provide little incentive for ,ervice to do internal wor*. 'uente "ills Toyota transfers
at mar*et prices because they want to measure each department as a stand-alone business and they want
to have the 9sed 6ar 8epartment get as much priority in the ,ervice area as any other customer.
The measurement focus in this business is on profit measures. 8o profit measures provide a good
indication of value creation in the car retailing business) The answer here is probably yes. This is
primarily a short-term business. 8ealerships are not ma*ing many investments that involve large lags
between time of investment and payoff /e.g., #:81, and they are not creating many intangible assets. The
one exception is customer goodwill, so it is not surprising that Toyota mandates that considerable effort
also be devoted to the measurement of 6,..
,ome useful class discussion can focus on the structure of the incentie plans. 8epending on the focus of
the class and the class time available, the instructor can have the students complete an incentive plan
matrix. +n a whiteboard, list any or all of the *ey incentive plan elements. (ith ade7uate time, . use the
following-
;. (ho is included in the plan)
2. <orm of payment
0. <re7uency of payment
=. >easure/s1 and their importance weighting
4. 'erformance standards
?. ,hape of reward@performance function
A. ,i%e of reward /expected, maximum1
8. <unding /e.g., is there a company bonus pool constraint)1
5. 9niformity /are people in the same role treated the same)1
.n an array of columns, list the roles for which the instructor wants to clarify the incentive system. These
might include the general manager, general sales manager, service manager, salespeople, service advisors,
service technicians.
Among other things, this analysis will show that-
The payments are all in cash
The payments are timely /monthly1
The bonuses are paid by formulaB there is no bonus pool constraint
There are no performance standards except in the service technician area where standards are set
by Toyota. .nternally set budgets are used to calibrate the payoff function, but goal-setting does
not seem important in this business.
The service manager reward@performance function is *in*ed upward to encourage beating the
budget.
The rewards are 7uite lucrative.
The assignment 7uestion as*s students for an evaluation of the system, which re7uires them to identify
good and bad points about it. .n addition to the points raised above, they might mention the following-
/-1 There are no deferred compensation elements that might provide retention and tax benefits.
/-1 This is an incentive system that would re7uire the company to pay si%able bonuses even if the
company is losing money.
/-1 There is no bonus for teamwor*. "ow much teamwor* is necessary) The case mentions one area
that could be improvedCservice referrals to sales.
/)1 .s the company paying too much) Answering this 7uestion would re7uire *nowledge of industry
benchmar*ing. ,ome of these data are provided in the case Appendices. 'uente "ills Toyota$s
practices seem in line with other dealerships.
/)1 The bonuses are all formula-based. (ould it be useful to allow some subjectivity in case
unforeseen events unfold)
2enerally, however, this measurement and incentive system probably must be considered as effective.
>uch of this system is dictated by Toyota, and this is an industry and company that has refined its
systems over the years. And within the Toyota family, 'uente "ills Toyota is a top performer.
Case 27-2: Axeon N.V.
Note: This case is unchanged from the Twelfth Edition.
Purpose of Case
The Axeon D.E. case was written to illustrate the effects of a management control system and the
supporting management processes on one specific, major decision in a decentrali%ed, multinational corpo-
ration. The situation, which is real, but disguised, illustrates the real world application of many manage-
ment accounting concepts, including incremental cost analysis, capital budgeting, sensitivity analysis, and
transfer pricing. 3ut perhaps more importantly, the case is about managing the cost@benefit tradeoffs that
are inherent in a decentrali%ed firm. The case illustrates some common management process problems in-
volving conflict between a parent and its foreign subsidiary. Eisible in the case are common attributes of
managerial behavior in a decentrali%ed organi%ation. The attributes are both positive /motivation, ini -
tiative1 and negative /suboptimi%ation, parochialism, indecisiveness1. 8ealing with these behaviors forces
students to consider issues about organi%ational design and control system administration.
3ecause of the breadth of issues raised, this is an excellent review case. >any students will become
emotionally engaged in the situation and, hence, the case becomes a good vehicle for discussing the
advantages and disadvantages of decentrali%ation and the problems faced in administering a control
system in a decentrali%ed environment.
Suggested Assignment Questions
;. .s construction of the new factory in the 9.F. in the best interest of Axeon D.E.)
2. .gnoring your answer to 7uestion ;, if the plant were not built and A#-=2 was shipped
from the Detherlands to the 9.F., what transfer price would be appropriate)
0. (hat should >r. van Geuven do)
Case Analysis
A. Product Sourcing Alternatives
8espite the existence of many relevant data in the case, the answer to the first assignment 7uestion is not
an easy one. There are two alternatives for producing A#-=2- 9.F. and Detherlands.
The best place to start is with the discounted cash flow analysis, a summary of which is presented in
&xhibit 2 of the case. 6ase &xhibits 0 to 4 provide bac*-up detail. The analysis is straightforward, and it
presents no computational difficulties. The new-factory project shows a positive net present value using
an 8H rate, which seems to be the marginal borrowing rate for "ollandsworth in the 9.F. And the project
shows a ;4H .##, which is above Axeon$s ;2H hurdle rate. ,o, based on these analyses, the project
loo*s worthwhile.
(as there anything wrong with this initial analysis) The analysis loo*s sophisticated, and most students
will instantly empathi%e with >r. (allingford. "owever, criticisms can be made of his analysis-
;. The end-of-project recovery value of e7uipment and wor*ing capital is a surrogate for
the cash flows after the end of seven years. .t is li*ely that this is biased downward. (hy would
all sales and profits stop after that date) .s there no learning that has come from this project)
8oes the project have no option value)
2. (allingford might well have presented some sensitivity analyses, reflecting more
pessimistic forecasts. These might have headed off some of the criticism that came later from
Axeon$s director of sales.
0. >ost importantly, although he appears to be ta*ing a corporate perspective and, hence,
comes across as an appealing character with whom most >3A students will identify, (allingford
is really parochial. "e did not consider the investment opportunity using a corporate-focused per-
spective. (allingford is indignant about Axeon managers$ rejection of his idea, but he is not as
innocent as he portrays himself. The main defect in his analysis was his failure to consider the
alternative of manufacturing in the Detherlands. The main missing information in his original
proposal was this implicit benchmark.
(allingford might also have proposed a two-step approach to selling A#-=2 in the 9.F. "e could
have demonstrated the si%e of the 9.F. mar*et using imports from the Detherlands, and then later
proposed to save the shipping and duty costs by manufacturing in the 9.F. .nstead, he chose the
more self-centered approach of going it alone from day ;. (hy didn$t (allingford present this
alternative) 'robably because he wanted to run his own plant /build his own empire1.
,o we see in the case that the 8utch managers argue that producing the A#-=2 in the Detherlands is an
even better alternative than that proposed by >r. (allingford. ,ome of their arguments 7uestion some of
the numbers in the "ollandsworth analysis. The Axeon director of sales />r. 8e #ijc*e1 7uestioned the
sales forecasts. 3ut who understands the 9.F. mar*et better, the local general manager or the functional
specialist in a foreign country)
The Axeon director of manufacturing />r. +osterling1 7uestioned both the 9.F. facility$s ability to
handle the manufacturing and the cost assumptions. ,tudents should thin* about his statements regarding
costs. The fixed costs are already sun*, so they are not relevant to the analysis. The analysis of the
variable costs /as shown in &xhibit ? of the case1 is misleading. As shown in &xhibit-TD ;, an average
variable cost of I;8?J on the full ;JJJ tons of production implies an incremental variable cost of I;8JJ
on the additional =JJ tons for 9.F. sales. This incremental variable cost is the relevant cost when com-
paring the two investment proposals. ,o >r. +osterling has overstated his case.
Dext, the instructor can turn to the discounted cash flow comparisons. &xhibit-TD 2 shows an analysis of
the Detherlands proposal. The revised analysis still favors the Detherlands proposal, mainly due to the
much greater initial investments for the 9.F. proposal.
9.F. 'roposal Detherlands 'roposal
Det 'resent Ealue K 8H I5;?,JJJ L;,288,A5J
.nternal #ate of #eturn 2JH 2?H
.n addition, ta*ing into account the higher ris*s involved in the 9.F. startup relative to the Detherlands
expansion, the 9.F. hurdle rate should be higher, which ma*es the Detherlands alternative even more
attractive.
+n the other hand, sensitivity analyses would show that the 9.F. managers could face significant sales
shortfalls and manufacturing problems and still have a project with a positive net present value. .n terms
of that criterion, or even the Axeon ;2H hurdle rate criterion, perhaps the impact of the startup
complications are not as significant as the Axeon functional managers believe them to be.
<urther, the irrelevance of the fixed costs in the Detherlands proposal ma*es sense only if the Detherlands
has, and will always have, excess capacity for producing A#-=2. .n other words, the assumption of %ero
opportunity cost of the fixed plant is valid only if the Detherlands capacity is greatly overbuilt. .f this is
the case, it suggests problems with Detherlands capacity planning in the past. The longer the time
hori%on, the better the 9.F. proposal loo*s.
+ther issues such as nationalism and perceived autonomy are also important. Axeon D.E. is increasingly
embracing a philosophy of decentrali%ation. Therefore, even if the 9.F. proposal is not optimal, forcing
manufacturing of A#-=2 in the Detherlands will undercut the perceived decentrali%ation in the company.
Another consideration is the nationalistic sentiments of the &nglish. This is suggested by the re7uirement
of a majority of local directors for the "ollandsworth board and the threatened resignation of >r. Dobel.
An increase in duties on A#-=2 is li*ely if the 9.F. viewed Axeon as exploiting the 9.F. mar*et from the
Detherlands.
<inally, >r. (allingford is *nown to be enthusiastic and innovative. .f he feels that he has less autonomy
than before and all the benefits go to the Detherlands, his enthusiasm and innovative spirit are bound to
suffer. "e might be less committed to developing the A#-=2 mar*et in the 9.F. And, based on his
outstanding performance at Axeon, opportunities for >r. (allingford to leave Axeon for a competitor are
probably abundant. 6an Axeon afford to lose him)
. !"y did #r. van $euven %e"ave as "e did&
.t is easy to be critical of >r. van Geuven in this case. "e caused much of the problem because he did not
manage the process very well. "e should learn that in some situations he should refrain from comment
when he is exposed to early-stage proposals at board meetings. "is early encouragement in this situation,
which he subse7uently reversed, caused much of the disappointment from the "ollandsworth personnel.
>r. van Geuven was also representing Axeon at the "ollandsworth board meeting. .deally, he should
have proposed the alternative of sourcing from the Detherlands early in the decision-ma*ing process.
>uch of >r. van Geuven$s behavior in the case was designed to find a way out of a problem that he
probably *new he had caused.
C. T"e Trans'er Pricing (ssue
.f the A#-=2 for the 9.F. is sourced from the 9.F., then the analysis is complete. "owever, if A#-=2 is
supplied from the Detherlands to the 9.F., then a transfer price must be established. There are a number
of transfer pricing possibilities-
;. !ar"et price. This would award most of the profit! to the producer, in this case, the
Detherlands.
2. Cost# or cost plus a mar"$up. 6ost can be defined anywhere from incremental
variable cost to full cost, and the mar*-up can be variable /a fixed amount per unit1 or a lump sum
/to compensate the Detherlands for costs of the increased wor*ing capital and the use of its
capacity1.
0. A two$boo" system. This would involve crediting the Detherlands at mar*et price but
charging the 9.F. only the variable cost. This would ma*e both divisions happy, presumably, but
it would lead to a double counting of profits and re7uire a corporate elimination.
.n many decentrali%ed companies, managers would allow the division managers to negotiate a fair
transfer price. 3ut in this case, because of the friction already generated between personnel in the 9.F.
and the Detherlands, negotiation may not be possible. ,o here >r. van Geuven probably has to choose a
transfer pricing method and a means of implementing his decision.
The total contribution per ton of A#-=2 is as follows /using figures from year two and following1-
,elling price L0,AJJ
Eariable costs /;,8JJ1
,hipping /;JJ1
8uty /;JJ1
6ontribution per ton L;,AJJ
&ach transfer pricing method allocates this contribution between the 9.F. and the Detherlands.
(hat transfer pricing@profit sharing arrangement is best) There is no correct solution. Transfer prices
provide one way of shaping managerial behavior. .f centrali%ed production is deemed to be the strategic
focus, then a transfer price giving more of the profits to the Detherlands will best signal this emphasis.
3ut if top management wants to reward local mar*et sensitivity and initiative, then a transfer price that
passes most of the profits on to the 9.F. is best. A thoughtful analysis of the transfer pricing issue will
lead to a discussion of broader strategic 7uestions.
). *+it, et-een Strategy. /rgani0ation Structure. and Control Syste1s
At this point in the class discussion, it is useful for the instructor to as* the following 7uestion- (hat is
Axeon$s corporate strategy) 6an you tell from the case what the *ey success factors are for Axeon)
The case does not provide much information about Axeon$s critical success factors. (hat is most critical
for AxeonCidentifying new applications and mar*ets or producing at low cost) The case does not
provide a clear answer to this 7uestion. "owever, one thing that is clear is the trend in Axeon$s interna-
tional expansion and in its organi%ational structure. E%hibit TN$& illustrates this trend. Axeon used to
have a centrali%ed, functional organi%ational structure, but the firm is moving to greater decentrali%ation
to ta*e advantage of geographical expertise in each of the areas in which it operates.
,hould the company$s critical success factor/s1 affect decisions about organi%ational structure and
management control systems) 6learly yes. This observation can lead naturally into a discussion of the
pros and cons of decentrali%ation vs. centrali%ation. 2enerally spea*ing, the advantages of centrali%ation
include economies of scale and technically s*illed employees. +n the other hand, decentrali%ation leads to
better relationships with local governments, sensitivity to the local mar*et, and local managers$ higher
perceived autonomy. E%hibit TN$& summari%es the trade-off between flexibility and efficiency as the
level of decentrali%ation increases. Two points are worth noting-
;1 >ar*et sensitivity is enhanced with increased decentrali%ation. Also, a decentral i%ed
manufacturing firm with a larger number of smaller facilities enables greater production flexibil-
ity.
21 &conomies of scale decrease with increased levels of decentrali%ation. This is because
a decentrali%ed organi%ation loses the cost savings from shared management overhead and larger,
more efficient plants.
.n summary, Axeon is facing two sets of pressures-
6entrali%ation ,trategy 8ecentrali%ation ,trategy
Fey success factor Gow cost Dew product applications
Fey function 'roduction >ar*eting
#esponsibility 6enters-
'roduction .nvestment center 6ost center /captive1
>ar*eting #evenue center 'rofit@.nvestment center
To some extent, Axeon$s control system is a compromise between full decentrali%ation, which would use
highly autonomous local investment centers, and full centrali%ation, which would have the local
organi%ations be only mar*eting outposts, revenue centers.
(e also see in the case that the current reward system includes a bonus based entirely on the local
division results. This seems to reinforce the decentrali%ation strategy. Alternatively, even under a decen-
trali%ation strategy, a bonus might be based partly on company-wide profits to promote cooperation
between divisional managers. A reward system can reinforce a chosen strategy as well as compensate for
the wea*nesses of the chosen strategy.
,o Axeon seems to be gradually embracing decentrali%ation. 3oth the line and the staff organi%ations
have significant power /decision rights1. The friction we see in the case is typical of a matrix organi %ation
where country managers li*e (allingford are often at odds with functional specialists at head7uarters.
>r. van Geuven has to play an integrator role. "e should smooth over the rough spots that will inevitably
arise.
2. !"at s"ould #r. van $euven do&
There is no correct solution hereB there are only pros and cons. The options are to source in the 9.F.,
source in the Detherlands, or delay. 8elay means possibly source in the 9.F. after the viability of the
mar*et has been demonstrated. As*ing students to vote as to their preference will reveal significant
differences in opinion. This vote, if one is ta*en, should come very late in the class.
(hile it appears that the economics of this decision favor the Detherlands, our own ta*e is that the
decision in this case should probably favor the 9.F. .f a decentrali%ed organi%ational structure is appro-
priate for Axeon, as it appears to be, then decentrali%ation must be allowed to wor*. >otivate the local
managers to ta*e initiative, and reward them for ta*ing that initiative. The "ollandsworth proposal
exceeds the established hurdles, so allow "ollandsworth managers to proceed with their plan. .f students
recommend the seemingly superior economic choice of sourcing the production, then they should just be
made to understand the negative motivational effects that choice will have on the motivation of the 9.F.
managers, and possibly even their successors.
Gonger-term, Axeon might try to implement a more sophisticated analytical procedure so that the right
alternatives are properly addressed in proposals such as that analy%ed here.
'hat happened(
.n the real world situation, the "ollandsworth project was rejected. The transfer price was set at L0AJJ, so
the full cost to "ollandsworth /including shipping and duty1 was L05JJ. .nitially "ollandsworth
managers tried to sell A#-=2 at a price of L=4JJ per ton, but the volume sold was miniscule. &ight
months later they lowered the price to L=2JJ. The annuali%ed rate of sales went to ;4J tons, and "ol-
landsworth managers thought that the potential was there for annual sales of 2JJ tons at that price. <our
months later they lowered the price to L0AJJ. The annuali%ed rate of sales went to 2AJ tons. (ith this
result, the "ollandsworth managers thought that they had clearly demonstrated the potential for annual
sales of =JJ tons at that price.
Axeon began an increased push to sell A#-=2 in other parts of the world. Muic*ly they filled the capacity
of the plant in the Detherlands. "ollandsworth eventually got permission to build a plant in the 9.F.
>r. (allingford did not leave the "ollandsworth immediately. .nstead, he focused on other issues and
continued to be a highly successful manager. ,ome years later, >r. van Geuven was promoted. >r.
(allingford was offered >r. van Geuven$s job, but he turned it down because he wanted his children to
be educated in the 9.F. ,oon after having made that decision, he was offered a position of managing
director of a larger 9.F. company, and he accepted that offer.
Student Ta"eaways
,tudents can learn a lot from this case. They can practice their technical s*ills, such as those re7uired to
do net present value analyses, and they can apply their transfer pricing theories to a real world-li*e
situation. 3ut perhaps most importantly, by seeing that what appears to be a good investment is not made
and understanding why not, students can see that the process of management is more intriguing and more
complicated than it usually appears in textboo*s.
Pedagogy
>any students see this case as difficult. ,ome amount of floundering with the numbers is useful, but if
the floundering continues, instructors should be prepared to step in and help clarify the analysis. That
understanding is essential before moving on to the more 7ualitative aspects of the case. ,uggested timing
in a A4-minute class is as follows-
! min. The plant investment decision
"# min. Transfer pricing alternatives and behavioral effects
"# min. The fit between the existing control system and the company$s strategic priorities
# min. (hat happened@summary
$# minutes
E%hibit TN$)
*eleant Cost Analysis
Calculation of incremental ariable cost per ton of manufacturing
A*$+, in the Netherlands for shipment to the -./.:
'rojected average variable cost in the Detherlands at ;JJJ tons I;,8?J
'rojected total variable cost in the Detherlands at ;JJJ tons I;,8?J,JJJ
'rojected average variable cost in the Detherlands at ?JJ tons I;,5JJ
'rojected total variable cost in the Detherlands at ?JJ tons I;,;=J,JJJ
Additional variable cost to produce =JJ tons for 9.F.
/I;,8?J,JJJ - I;,;=J,JJJ1 IA2J,JJJ
.ncremental variable cost to produce =JJ tons for 9.F.
/IA2J,JJJ @ =JJ1 I;,8JJ
*eised ariable cost per ton 0Different from E%hibit 12:
.ncremental manufacturing variable cost I;,8JJ
,hipping from Detherlands to 9.F. ;JJ
9.F. import duty ;JJ
Total variable cost per ton I2,JJJ
Total variable cost, =JJ tons to 9.F. I8JJ,JJJ
E%hibit TN$,
Discounted Cash 3low Analysis 4 Netherlands Proposal
/<igures in rows /21-/=1 in IB row /;1 and rows /?1-/51 in IJJJ1
%ear ! " & ' # ( $
/;1
(or*ing capital
-;2J -2J -2J ;?J
/21
,ales price
per ton
=,JJJ 0,AJJ 0,AJJ 0,AJJ 0,AJJ 0,AJJ 0,AJJ
/01
.ncremental cost per ton
2,JJJ 2,JJJ 2,JJJ 2,JJJ 2,JJJ 2,JJJ 2,JJJ
/=1
6ontribution per ton
/21 N /01
2,JJJ ;,AJJ ;,AJJ ;,AJJ ;,AJJ ;,AJJ ;,AJJ
/41
,ales
/in tons1
2JJ 0JJ =JJ =JJ =JJ =JJ =JJ
/?1
Total contribution
/=1 x /41
=JJ 4;J ?8J ?8J ?8J ?8J ?8J
/A1 'romotion costs 2?J ;4J ;JJ ;JJ ;JJ ;JJ ;JJ
/81
Tax
=JH of
/?1 N /A1
4? ;== 202 202 202 202 202
/51
Det cash flow after tax
/;1O/?1-/A1-/81 -;2J ?= ;5? 0=8 0=8 0=8 0=8 4J8
Det 'resent Ealue at 8H I;,288,A5J
'aybac* 'eriod ;P years
.nternal #ate of #eturn 2?H
E%hibit TN$&
3le%ibility and Efficiency as a 3unction of the 5eel of Decentrali6ation
Economies of scale7
Production efficiency
3le%ibility7
!ar"et sensitiity
A%eon