escalation_of_commitment_measures

Escalation of commitment outcomes measures

From The Psychology of Sunk Cost (Arkes & Blumer, 1985, OBHDP):

Assume that you have spent $100 on a ticket for a weekend ski trip to Michigan. Several weeks later you buy a $50 ticket for a weekend ski trip to Wisconsin. You think you will enjoy the Wisconsin ski trip more than the Michigan ski trip. As you are putting your just-purchased Wisconsin ski trip ticket in your wallet, you notice that the Michigan ski trip and the Wisconsin ski trip are for the same weekend! It’s too late to sell either ticket, and you cannot return either one. You must use one ticket and not the other. Which ski trip will you go on?

  • $100 ski trip to Michigan
  • $50 ski trip to Wisconsin

From Throwing Good Money After Bad: The Effect of Sunk Costs on theDecision to Escalate Commitment to an Ongoing Project (Garland, 1990, JAP):

You are the President of Aero-Flite Corporation, an airplane manufacturer. You have spent million dollars of the 10 mil-lion dollars budgeted for a research project to develop a radar scrambling device that would render a plane undetectable by conventional radar (in effect a radar-blank plane). The project is % complete. Another firm has begun marketing a similar device that takes up much less space and is much easier to operate than Aero-Flite's.

  • How likely is it that, if faced with this situation, you personally would decide to use the last X million dollars to complete this project?“
  • How likely is it that, if faced with this situation, you personally would authorize the next X million dollars to continue with the project?”
  • How likely is it that if you decided to complete this project using the last million dollars of your budget, your company would realize a profit?

From The Role of Anticipated Regret in Escalation of Commitment (Wong & Kwong, 2007, JAP, the Appendix:

The version used in the responsible/high-regret possibility condi-tion was as follows, with the wording used in the low regret possi-bility condition in parentheses (1 Hong Kong dollar13 U.S. cents).

One day, you and two other friends, Peter and Ken, were on the wayto take a bus. Unfortunately, you saw that the bus you wanted to takeleft the bus-stop right before you arrived. You missed the bus andneeded to wait for the next bus. You remembered that the schedule ofthis bus was quite frequent and said, “This is a frequent bus. We won’twait too long. The next bus will come in around 10 minutes.”Although Ken agreed with you, Peter expressed doubt. Peter said, “Iremember that the schedule of this bus is very infrequent. Last time Ispent 40 minutes waiting here for the next bus.” You and Peterengaged in a debate about how frequent this bus is.Accordingly, you both agreed to have a bet. You agreed to wait 15minutes at the bus-stop. You would then pay HK$10 to Peter for everyminute in this 15-minute period until the bus came. Thus, if the busdid not come within 15 minutes, you would lose HK$150 in total toPeter. However, if the bus came within 15 minutes, he would payHK$150 to you, though he would not return any money he had gainedin the previous minutes.Ten minutes later, the bus had not yet come, and so you had paidHK$100 to Peter. Now, Peter suggested that you could stop the bet butstay for five more minutes (and then leave the bus-stop to take thetrain). You would then (would not) learn if the bus came in the next5 minutes. If you accepted Peter’s suggestion and terminated the bet,Peter would not pay HK$150 to you and you would not have the riskof losing an additional HK$50 to Peter. Alternatively, you couldchoose to continue the bet and wait 5 minutes more. Doing so, youwould risk losing an additional HK$50 to Peter although there was achance that Peter would pay HK$150 to you.

The version used in the personally nonresponsible condition wasidentical to the above one, except that (a) the first bet was betweenKen and Peter, and Ken paid $100 to Peter, and (b) Ken passed thesecond bet to the participant. Thus, participants were not respon-sible for the former loss of $100, though there was negativefeedback on their belief that the bus would come soon. Afterreading the scenario, participants were asked to indicate theirwillingness to continue the bet by giving a probability ratingranging between 0 (absolutely no) and 100 (absolutely yes).

Another version used in Study 2b, perhaps better use this one:

The version used in the personally responsible condition was asfollows.

One day, you and two other friends, Peter and Ken, were on the wayto take a bus. Unfortunately, you saw that the bus you wanted to takeleft the bus-stop right before you arrived. You missed the bus andneeded to wait for the next bus. You remembered that the schedule ofthis bus was quite frequent and said, “This is a frequent bus. We won’twait too long. The next bus will come within 10 minutes.” AlthoughKen agreed with you, Peter expressed doubt. Peter said, “I rememberthat the schedule of this bus is very infrequent. Last time, I spent 40minutes waiting here for the next bus.” You and Peter engaged in adebate about how frequent this bus was.Accordingly, you both agreed to have a bet. If the next bus arrivedwithin 15 minutes, Peter would pay HK$100 to you. However, if therewas no bus arriving within 15 minutes, you would pay HK$100 toPeter.Fifteen minutes passed and no bus came. You lost HK$100 to Peter.Then, Peter proposed another bet. He suggested waiting for the bus 15minutes longer. If it arrived within 15 minutes, he would return theHK$100 back to you. However, if there was no bus arriving in thenext 15 minutes, you would pay another HK$100 to Peter.

The version used in the personally nonresponsible condition wasidentical to the above one, except that (a) the first bet was betweenKen and Peter, and Ken paid HK$100 to Peter, and (b) Ken passedthe second bet to the participant. Thus, participants were notresponsible for the former loss of HK$100, though there was anegative feedback on their belief that the bus would come soon.Escalation of commitment was indicated by participants’ ratingson the willingness to continue the bet, which ranged from 0(absolutely not willing) to 100 (absolutely willing)

From The Role of Anticipated Regret in Escalation of Commitment (Wong & Kwong, 2007, JAP, the Appendix:

The version used in the personally responsible/high-regret pos-sibility condition was as follows, with the wording used in thepersonally nonresponsible condition in parentheses ( ) and the wording used in the low-regret possibility condition in brackets [ ].(1 Hong Kong dollar13 U.S. cents).

Last year you (Frank, your colleague) had a plan to organize a jazz-dancing program in school and would apply for the Quality Education Fund (QEF), this year for its establishment. You (He) have(has) given many efforts in preparation, including information collection, co-ordination, an attitude survey within school, and meetings with colleagues and students, etc. This year you continue to follow-upthe jazz-dancing program (Frank retired last year and this year you were assigned to follow-up the jazz-dancing program).The Education Department would evaluate all QEF proposals on a10-point scale, with 1 representing poor, 5 representing average, and 10 representing excellent. According to your past experience, you expectedthat this proposal would be rated 7, which was higher than the past threshold of accepting a QEF proposal on a jazz-dancing program, i.e.,the threshold was 6. Your estimates have been very accurate.Right before you began to write the proposal, the QEF officer gave two public announcements. First, it had funded too many jazz-dancing programs in the last two years, e.g., over 150 programs, and further funding to similar programs would be highly selective this year.Second, to let teachers understand the new standard [to avoid teachers over-interpreting the results of this particular year], they will [NOT]disclose to the public the minimum rating of accepting a jazz-dancingprogram proposal on the basis of the accepted proposal in this year.That is, you will [NOT] learn whether or not your estimate of your proposal rating will be higher than, at, or lower than the new standard.The decision you need to make now is to either abandon the jazz-dancing proposal and write another proposal for another dancing program that isless selectively funded or continue writing the same proposal.

Respondents gave a rating ranging between 0 (absolutely no)and 100 (absolutely yes) to indicate the extent to which they wouldbe willing to continue to write the current proposal. We used this willingness rating to indicate escalation of commitment.

from De-Escalation of Commitment in Oil Exploration:When Sunk Costs and Negative Feedback Coincide (Garland, Sandefur, & Rogers, 1990, JAP):

The decision scenario was designed to portray, albeit in simplifiedform, a realistic petroleum drilling situation. All information pre-sented was easily understandable to petroleum geologists. To examinethe impact of dry wells on the decision to drill additional wells, allother things being equal, we deliberately excluded from our scenariosthe typically rich geological and geophysical data that geologists re-ceive when they drill a dry well. The actual scenario used in the one-dry-well condition was as follows:

Your company has acquired a good sized lease position within awell known North American basin, the position was acquired at acost of $135,000.The company has recently drilled a 50 bopd [barrels of oil perday] discovery well on the lease. The acreage position is such thata minimum of five more wells can be drilled. The drilling part-nership provides a 1 million dollar budget for the five subsequentwells. The budgeted cost per well is $200,000 ($100,000 drillingcost; $100,000 completion cost). You have been given the finalauthority to authorize all expenditures on this project. The discov-ery well has proven out both your original exploration approachand associated geophysical data. However, the confirmation wellwas a $100,000 dry hole. Your total cost for this nonproducing wellis, thus, $100,000

Subjects in the two-, three- and four-dry-well conditions read thesame scenario, except that it was specified that the confirmation welland the next one, two, or three wells had all been dry holes. In additionit was specified that the total cost for these two, three, or four dry holeswas $200,000, $300,000, or $400,000. By emphasizing the total cost ofdry wells in each condition, we hoped to make sunk costs salient to allsubjects.A number of points should be made about how practicing petroleumgeologists would interpret the facts presented to them in these sce-narios. First, the positive information provided about the discoverywell clearly implies that there is oil on the land. Second, it would beassumed that the geologist had made the specific choice as to the bestlocation to drill all wells on the land. Third, the relatively low cost ofthe land, compared with the cost of drilling, is typical in the industry.Finally, it is typical to authorize both the cost of drilling and comple-tion (i.e., actually extracting the oil should viable commercial quanti-ties be found) for each well, even though the actual risk involved is onlythe dry-well cost, which in the present case is $100,000.

Immediately following the scenario were two questions.

  1. The firstquestion asked subjects to indicate “on a scale from 0-100 how Iikely itis that if faced with this situation, you would authorize another$200,000 to drill the next [or third, fourth, last, depending on thesubject's condition] well in the program?” Subjects responded by cir-cling a point along a 100-point scale marked definitely would not autho-rize (1) and definitely would authorize (100) at the endpoints. The mid-point of the scale was marked even chance.
  2. The second question asked, “Regardless of how you answered theprevious question, what is your perception of the likelihood that thenext well to be drilled on this prospect would produce 50 or morebopd?” The choice of 50 bopd was based on an expert judgment that,given the budget for this project and the price per barrel for oil at thetime this research was conducted, this was about the smallest yieldthat would produce a reasonable return on investment. Subjects againresponded along a 100-point scale marked definitely would not produce50 or more bopd (1) and definitely would produce 50 or more bopd (100)at the endpoints and even chance at the midpoint

Heath, 1995, OBHDP: it's complicated to copy, download and read.

  • escalation_of_commitment_measures.txt
  • Last modified: 2016/03/28 21:28
  • by filination