Discussion: View Thread

Chiara:

Among the serious problems related to performance appraisal, one of the most deleterious is that they are often given once a year, thereby  subjecting the appraiser to all sorts of errors in judgment and decision-making (e.g., recency, memory recall, etc.).  Another problem is the lack of a standard for the appraisals.  From your note I couldn't tell if these issues are characteristic of your situation, but, if they are, a suggestion I have is to begin with your organization's over-arching strategy for the year, translate that into organizational goals, and cascade these goals down throughout, with each department in the hierarchy having to determine what goals they have to achieve in order to contribute to the goals of their reporting department.  From this individual goals are derived.  This kind of process entails frequent discussion and review throughout the organization, the expected result being a much more efficient organization because there is "rhyme and reason" to the appraisal process. 

Bernard Liebowitz, PhD, CMC

Liebowitz & Associates, PC

980 No. <st1:street w:st="on"><st1:address w:st="on">Michigan Ave.</st1:address></st1:street>, <st1:address w:st="on"><st1:street w:st="on">Ste.</st1:street> 1400</st1:address>

<st1:place w:st="on"><st1:city w:st="on">Chicago</st1:city>, <st1:state w:st="on">Il</st1:state>   <st1:postalcode w:st="on">60611</st1:postalcode></st1:place>

tel}   312/214-3583   fax}  312/214-3510

email}  bernie@liebowitzassoc.com

web}    www.liebowitzassoc.com

From: Chiara Pollina
Date: Fri, 22 May 2009 16:27:06 +0200
To: <OB@AOMLISTS.PACE.EDU>
Subject: [OB-LIST] Pre-defined performance appraisal results

Dear Colleagues,

in <st1:country-region w:st="on"><st1:place w:st="on">Italy</st1:place></st1:country-region> the result of performance appraisal for public manager tend to be very high, even if the real results are poor.

Our minister for pubblic administration is going to adopt a law that says that 25% of people must have a performance appraisal "high" (100% of bonus); 50% must have an appraisal "medium" (50% of bonus) and 25% must have an appraisal "low" (100% of bonus).

I think that such solution has a lot of cons:

- it will cut out all OCB

- this way it will reduce organization performance

- it discriminate in a too drastic way, creating only three classes of performance.

What do you think about, and what kind of solution do you think could be used in order to solve the problem?

Thanks a lot,

Chiara Pollina

Chiara,

Here is an article in BusinessWeek about forced distribution performance appraisal.  I make my HR students read it every year; it talks about some positives and negatives of this approach, as well as what some companies are doing.

http://www.businessweek.com/magazine/content/06_02/b3966060.htm?chan=search

In terms of academic research on the topic, here are some references:

Scullen, S.E., Bergey, P.K., & Aiman-Smith, L. (2005). Forced distribution rating systems and the improvement of workforce potential: A baseline simulation.  Personnel Psychology, 58: 1-32.

Roch, S.G., Sternburgh, A.M., & Caputo, P.M. (2007). Absolute vs. relative performance rating formats: Implications for fairness and organizational justice.  International Journal of Selection and Assessment, 15: 302-316.

Blume, B.D., Baldwin, T.T., & Rubin, R.S. (2009). Reactions to Different Types of Forced Distribution Performance Evaluation Systems.  Journal of Business and Psychology, 24: 77-91.

Good luck!

Laura

I would also add that forced distributions are done to make the job of the person requiring them easier.  My opinion is that we should be looking to find a way that is fair to those being ranked and this is harder.  But then that is why the people requiring the rankings are paid the big bucks.  Anybody can do a forced ranking but only the truly professional manager will find a way to do this in a way that truly evaluates the performance of others.  No matter how you cut it forced rankings still have a lot of "friendship factor" in them:  "I like Mary and she's a good person".

I don't have it handy but you might search the Wall Street Journal of within the last six months.  It had a terrific article. 

Buena Suerte y si fuera tan amable, dejanos (o por lo menos su servidor) saber de lo que decide.

G.

Gary D. Robinson, Ph.D.

Chair, General Business

School of Business and Technology

Telephone: 1-888-Call Capella (227-3552) Ext. 4641 (Pacific Time)

Telephone Direct: 206-232-7980

Mobile: 612-719-6954

Fax: 206-236-2709

Gary.Robinson@Capella.edu

docrobinson@comcast.net

www.capella.edu

Skype: docrobinson

Chiara, what you are describing is called a "forced distribution system" and is not uncommon in organizations.  Prior to my academic career, I worked for Intel.  Intel uses a forced system with quartiles.  I would not necessarily predict any of you are anticipating from this system, but there are definitely some down sides.  

First, is sounds like your current system does not do a good job of discriminating anyway, so this would not seem to be necessarily a negative to me.  However, it sounds like your system does not employ, or at least enforce the use of effective criteria.  Any system is only going to be as good as the criteria that is used, and the accuracy of recording actual performance.  A forced distribution system can work well if the criteria are relevant, comprehensive and well communicated ahead of time, and if there are appropriate ways to measure or at least estimate accurately the actual performance.

Where forced distribution systems really do poorly is when you have either little variance within a group or comparisons across groups or both.  For example, group A may be a set of highly motivated, talented individuals.  The marginal difference between the top person and the bottom person may be very small, which this system would not catch.  Group B may have a few talented individuals with a larger spread between the rest.  The top quartile of group A and the top quartile of group B contain somewhat comparable employees.   But the middle quartiles are not similar at all.  As you can see, forced ranking can obscure actual performance when you need to make decisions about someone's true contribution to the company.  And, what you can end up with is someone who is performing well and contributing to the organization and yet is in the bottom quartile or someone who is not contributing much but is in the top quartile.  And you can end up with both in the same organization.

So, I would say that the system that your administration is proposing is not necessarily bad and has many positive aspects, but only if it is implemented as part of a comprehensive approach that is well thought out.  Just adding this rule to your existing system is likely to result in issues of both procedural and distributive justice.  You may want to look to other organizations (like Intel) that have successfully (more or less ;-) implemented these systems.

Jeff Peterson
Utah Valley University


On 5/22/09 7:27 AM, "Chiara Pollina" <chiara.pollina@GMAIL.COM> wrote:

Dear Colleagues,
in Italy the result of performance appraisal for public manager tend to be very high, even if the real results are poor.
Our minister for pubblic administration is going to adopt a law that says that 25% of people must have a performance appraisal "high" (100% of bonus); 50% must have an appraisal "medium" (50% of bonus) and 25% must have an appraisal "low" (100% of bonus).
I think that such solution has a lot of cons:
- it will cut out all OCB
- this way it will reduce organization performance
- it discriminate in a too drastic way, creating only three classes of performance.
What do you think about, and what kind of solution do you think could be used in order to solve the problem?
Thanks a lot,
Chiara Pollina

I would also add that forced distributions are done to make the job of the person requiring them easier.  My opinion is that we should be looking to find a way that is fair to those being ranked and this is harder.  But then that is why the people requiring the rankings are paid the big bucks.  Anybody can do a forced ranking but only the truly professional manager will find a way to do this in a way that truly evaluates the performance of others.  No matter how you cut it forced rankings still have a lot of "friendship factor" in them:  "I like Mary and she's a good person".

I don't have it handy but you might search the Wall Street Journal of within the last six months.  It had a terrific article. 

Buena Suerte y si fuera tan amable, dejanos (o por lo menos su servidor) saber de lo que decide.

G.

Gary D. Robinson, Ph.D.

Chair, General Business

School of Business and Technology

Telephone: 1-888-Call Capella (227-3552) Ext. 4641 (Pacific Time)

Telephone Direct: 206-232-7980

Mobile: 612-719-6954

Fax: 206-236-2709

Gary.Robinson@Capella.edu

docrobinson@comcast.net

www.capella.edu

Skype: docrobinson

Chiara, what you are describing is called a "forced distribution system" and is not uncommon in organizations.  Prior to my academic career, I worked for Intel.  Intel uses a forced system with quartiles.  I would not necessarily predict any of you are anticipating from this system, but there are definitely some down sides.  

First, is sounds like your current system does not do a good job of discriminating anyway, so this would not seem to be necessarily a negative to me.  However, it sounds like your system does not employ, or at least enforce the use of effective criteria.  Any system is only going to be as good as the criteria that is used, and the accuracy of recording actual performance.  A forced distribution system can work well if the criteria are relevant, comprehensive and well communicated ahead of time, and if there are appropriate ways to measure or at least estimate accurately the actual performance.

Where forced distribution systems really do poorly is when you have either little variance within a group or comparisons across groups or both.  For example, group A may be a set of highly motivated, talented individuals.  The marginal difference between the top person and the bottom person may be very small, which this system would not catch.  Group B may have a few talented individuals with a larger spread between the rest.  The top quartile of group A and the top quartile of group B contain somewhat comparable employees.   But the middle quartiles are not similar at all.  As you can see, forced ranking can obscure actual performance when you need to make decisions about someone's true contribution to the company.  And, what you can end up with is someone who is performing well and contributing to the organization and yet is in the bottom quartile or someone who is not contributing much but is in the top quartile.  And you can end up with both in the same organization.

So, I would say that the system that your administration is proposing is not necessarily bad and has many positive aspects, but only if it is implemented as part of a comprehensive approach that is well thought out.  Just adding this rule to your existing system is likely to result in issues of both procedural and distributive justice.  You may want to look to other organizations (like Intel) that have successfully (more or less ;-) implemented these systems.

Jeff Peterson
Utah Valley University


On 5/22/09 7:27 AM, "Chiara Pollina" <chiara.pollina@GMAIL.COM> wrote:

Dear Colleagues,
in Italy the result of performance appraisal for public manager tend to be very high, even if the real results are poor.
Our minister for pubblic administration is going to adopt a law that says that 25% of people must have a performance appraisal "high" (100% of bonus); 50% must have an appraisal "medium" (50% of bonus) and 25% must have an appraisal "low" (100% of bonus).
I think that such solution has a lot of cons:
- it will cut out all OCB
- this way it will reduce organization performance
- it discriminate in a too drastic way, creating only three classes of performance.
What do you think about, and what kind of solution do you think could be used in order to solve the problem?
Thanks a lot,
Chiara Pollina

Chiara (and all):

In addition to consideration of the pros and cons of the "forced ranking" approach, you may also want to consider literature showing benefits (e.g., validity; rating accuracy; more natural for raters) arising from appraisals using "relative" (explicit social comparison) vs. more traditional "absolute" standards of comparison. Forced distributions and straight rankings are not the only (or necessarily the best) ways to solicit relative appraisals.. The following article will contains references to other relevant literature.

Goffin, R. D., Jelley, R. B., Powell, D. M., & Johnston, N. G. (2009). Facilitating social comparisons in performance appraisal: The Relative Percentile Method. Human Resource Management, 48(2), 251-267.

Cheers,

Blake

Chiara,

Here is an article in BusinessWeek about forced distribution performance appraisal.  I make my HR students read it every year; it talks about some positives and negatives of this approach, as well as what some companies are doing.

http://www.businessweek.com/magazine/content/06_02/b3966060.htm?chan=search

In terms of academic research on the topic, here are some references:

Scullen, S.E., Bergey, P.K., & Aiman-Smith, L. (2005). Forced distribution rating systems and the improvement of workforce potential: A baseline simulation.  Personnel Psychology, 58: 1-32.

Roch, S.G., Sternburgh, A.M., & Caputo, P.M. (2007). Absolute vs. relative performance rating formats: Implications for fairness and organizational justice.  International Journal of Selection and Assessment, 15: 302-316.

Blume, B.D., Baldwin, T.T., & Rubin, R.S. (2009). Reactions to Different Types of Forced Distribution Performance Evaluation Systems.  Journal of Business and Psychology, 24: 77-91.

Good luck!

Laura