22 MAART 2019
ASSIGNMENT 3: ANALYSIS AND
DISCUSSION OF RESULTS OF BEER GAME
INTERNATIONAL BUSINESS & SUPPLY CHAIN MARKETING
, The bullwhip effect in the beer game
The bullwhip effect is the phenomenon where orders to the supplier tend to have larger variance than
sales to the buyer (i.e., demand distortion), and the distortion propagates upstream in an amplified
form (i.e., variance amplification) (Lee, et al., 1997). The beer game played consisted of a multi-
period inventory model where demand was non-stationary over time, so in each week demand was
different for the retailer and consequently for the other members in the supply chain. The effect in
our game, as shown in Graph 1 of the Appendix, shows the bullwhip effect. As can be seen, the orders
placed by the retailer, which represent the actual demand by the end customer, are less variable than
the orders placed by the other players. Especially the wholesaler has experienced very high variability
in, for example, week 11 to 12. The variance amplification can be seen at the distributor and hereafter
the factory level, where this variance in requested orders takes place about two weeks later due to the
delay present in the game.
Causes of the bullwhip effect present in the beer game
When discussing the causes of the bullwhip effect we can distinguish two groups, which are
operational and behavioural causes.
Within the operational causes there are four types of causes namely demand signal processing, the
rationing game, order batching, and price variations (Lee, et al., 1997). The first two causes can be
seen as internal ones since they appear because each member of the supply chain tries to optimize its
own inventory, while the second two can be seen as external ones since they are reactions to market
dynamics (Lee, et al., 1997). Demand signal processing can occur when demand is non-stationary
and it means that supply chain members use demand information from past orders to update their
order forecasts (Lee, et al., 1997). To put it in another way, members place orders based on the
demand of previous periods so if demand was large in the past, then they place a larger order next
time. This can result in a big fluctuation within the supply chain because the members solely focus
on order information and lose track of real demand (Lee, et al., 1997). The rationing game occurs
when there is a shortage of supply and retailers order more than they would if the supply was
unlimited because they want to get the largest share of the products from the manufacturer or
wholesaler and this creates a higher variability in demand which leads to the bullwhip effect (Lee, et
al., 1997). Order batching happens when there are fixed order costs and because of this members of
the supply chain accumulate their demand before ordering instead of ordering in every period (Lee,
et al., 1997). The company that is subject to periodic ordering is exposed to a high variability in
demand and because of this order batching leads to the bullwhip effect (Moll, 2013). Price variations
means that the price of the product changes over time and because of this sometimes when the price
is very favourable and at the same time inventory costs are low it is worth for members to buy more
products at once than they can actually sell and because of the big difference between the purchased
and the sold quantities this results in the bullwhip effect (Moll, 2013).
If we look at the other group of causes namely the behavioural causes the literature mentions two
main sources of the bullwhip effect: bounded rationality and overreaction. Because human beings do
not behave rationally in every case even if all the operational causes are removed and the demand is
known by every member, there is a decision bias since managers are likely to underweight the supply
line which can lead to the bullwhip effect (Croson & Donohue, 2006). Furthermore, people in charge
of demand management tend to overreact when there is an increase or a decrease in orders from their
immediate upstream neighbours (Forrester, 1961). Overreaction can be responsible for causing the
bullwhip effect alone, however, it can be further increased by the transmission of information and
materials (Forrester, 1961).
1