I. Recruitment

1
I. Recruitment
2
Hiring the Right People

• Example 1 Jimbo Corp. has several openings for
  new managers. After advertising in the paper,
  the HR Department takes in 400 applications for
  every position. Is there anything that they can
  do in the future to reduce the load of looking at
  all of these applications?
• Example 2 Lisa Corp. has numerous openings for
  individuals to do cold calling. The HR Dept.
  has a hard time filling the jobs. There is a
  high turnover rate. It is suggested that they
  should not describe the job as selling so that
  they can get more applicants. What should they
  do?

3
1. Recruiting Strategies

• In this section we develop a simple model
• get used to our approach
• start you thinking like an economist
• You have a job to fill, would like to hire the
  best person. How can you do this?
• One strategy is to offer high pay, hoping to
  attract high skilled-applicants
• This will succeed, but at a cost high pay!
• there is an additional cost of a simple high pay
  strategy (see below)
• b/c of these problems, most firms sort applicants

4
Credentials

• One approach to sorting hire applicants with
  specific credentials
• e.g., specific items on a resume specific work
  experience or education
• Generally, people who have credentials will be
  higher priced in the labor market. How do you
  decide if a credential is worth paying for?

5
How to Decide on requiring a credential

• The key is to compare productivity to the wage.
  If the RATIO of wage to productivity is lower for
  high quality workers than low quality workers,
  use high quality!
• LABOR RATES AND LABOR COSTS ARE NOT THE SAME
  THING
• Usually measured as labor cost/unit of output ,
  i.e. the cost of an additional unit of output
  should be kept as low as possible. If
  individuals who have the credential have lower
  unit labor costs they should be hired even though
  their wages are higher.

6
Credentials Continued

• What is a useful credential for hiring?
• ability to perform at obtaining the credential
  must be positively correlated with ability to
  perform the job (the more so, the better)
• obtaining the credential must be relatively
  cheaper for better-qualified workers than for
  worse-qualified (Why is this important?)
• When are high credentials likely to be more
  important?
• The role of capital intensity
• The role of the design of jobs
• Independent jobs
• Team work

7
Credentials Continued

• What have been shown to be the most important
  credentials?
• Education
• Experience
• Analytic ability
• How does the difference in labor costs and labor
  rates wage/hour) explain why the U.S. can have
  free trade without high unemployment?
• But credentials are imperfect
• they are only imperfectly correlated with job
  performance
• some elements of job performance are intangible,
  hard to measure based only on a resume
• thus, some further screening is required

8
Why Dont Interviews Work Well?

• Interviewers often have the wrong incentives when
  they are judged by the number of persons who
  accept the job offer
• What might be better?

9
2. Screening

• You chose a set of candidates from resumes,
  recommendations, etc. What next?
• you can either take your chances, or you can
  screen them further
• Assumptions
• two worker types, a b. p probability worker
  is type a
• W wage Q productivity Qa gt W gt Qb
• Pra profit from hiring a QaW gt 0 Lb loss
  from hiring b QbW lt 0
• you can screen applicants at cost s, before
  hiring, with accuracy q

What hidden assumption are we making here about
the likelihood that each type applies for the job?
10
How Much Screening?

• Profit from hiring an applicant without screening
• ?dont screen pPra (1-p)Lb
• Profit from hiring an applicant only after
  screening
• ?screen pqPra (1-p)(1-q)Lb s
• The benefit from screening is the difference,
• D? p(1-q)Pra q(1-p)Lb s

11
How Much Screening?

• D? p(1-q)Pra q(1-p)Lb s
• Screening is more profitable when
• the test is more effective
• screening costs s are small (dD/ds lt 0)
• test is more accurate (dD/dq gt 0)
• test is more discriminating smaller fraction p
  employed after screening (dD/dp lt 0)
• the stakes are higher
• wage W is higher (dD/dW gt 0)
• employing those targeted by screening would be
  more costly (dD/dLb lt 0)
• note absolute value of L in the line above, as L
  lt 0 (EXAMPLES?)
• longer period the hired employee is likely to
  stay with the firm
• intuitively but this is complicated see below
  for more
• SO WHEN IS INTENSIVE SCREENING LIKELY TO BE
  EMPLOYED?

12
Who Pays For, Benefits From, Screening?

• Results of the screening, or the fact that the
  firm screens carefully, are often public
  information
• must pay at least W Q average workers
  productivity
• screening increases productivity, but also
  increases the wage in equilibrium
• but then D? s lt 0!
• for this to work, the applicant must pay for the
  screen hope to pass. How?
• going to school to earn a degree or credential
• accepting lower pay during probation
• The screen may be viewed as
• an investment by the worker
• education is an important pre-market screen
• a service sold by the firm (or school) to the
  worker (or student)
• e.g., internships entry-level jobs in
  professional service firms

13
Probation

• Studies suggest pre-job screening (interviews,
  psychometric testing) is far from perfect (when
  is it likely to be better?)
• the next step on-the-job screening should make
  accuracy q higher
• pay lower W during probation, higher W after for
  those retained
• the difference in pay must be high enough to
  attract those w/ high skills
• the probability of weeding out those w/ low
  skills must be high enough
• Turnover costs reduce the benefit from probation
• e.g., in Italy, if firing is found to be without
  cause, the firm must
• take the employee back
• pay lost wages plus social insurance
  contributions
• pay a fine to social security system up to 200
  of original amount due
• but is allowed to fire at will during first 3
  months of employment
• in Indonesia, if an employee is fired or quits,
  the firm must pay 2 months of severance for every
  year the employee worked for the firm

14
Probation (Continued)

• Public Sector in the U.S. Civil Service laws
  make firing difficult after probation
• Unionized Sector Required arbitration often
  makes it difficult to fire a worker
• At Will doctrine in U.S. becoming eroded by
  state statutes and judicial rulings

15
Temps

• Where termination is costly or difficult,
  consider temps
• let temp agency do screening
• pseudo probation during temporary employment
• In practice
• employers often permanently hire some temps from
  agency
• often for a fee or expectation of future business
• some temp agencies explicitly market themselves
  as placement agencies
• use is higher where laws against firing are
  tougher (across states, countries)
• growing in European labor markets (e.g., Spain)
• governments (e.g., France) making it easier to
  hire temps
• a similar possibility hire as a contract worker
  initially

16
3. Signaling

• Not everyone invests in credentials. Why not?
• individuals have private information about the
  odds of passing the screen
• Suppose now that workers know if they are type a
  or b
• firm could offer W Qa, not hire type bs
• whats wrong with that scenario?

• Now we have to worry about adverse selection
• I don't want to belong to any club that will
  accept me as a member
• Groucho Marx

17
Inducing Self Selection
We assumed p 1p were fixed, but they are based
on choices made by potential applicants. Can we
affect their choices, changing their probability
of applying?

• This was our approach above. Can we structure the
  offer to induce self selection?
• we want to induce a-types to join b-types to not
  join
• types a b as before
• assume the best alternatives they could earn
  elsewhere are Wb lt Wa
• 2 periods ignore discounting
• accuracy q of probation screening

18
Motivating Self Selection
Applicant considers NPV of job offer, decides
whether or not to apply (so probabilities change)
Probation
Post-probation
19
Probation Again

• Offer W1 in year 1 W2 in year 2 only to those
  retained
• a-types should expect better than elsewhere Þ
  W1qW2(1-q)Wa gt 2Wa
• b-types should expect worse than elsewhere Þ
  W1qWb(1-q)W2 lt 2Wb
• A little algebra shows that this will work if
• W2 gt Wa (WaW1)/q gt Wa
• W1 lt Wb (1-q)(WbW2) lt Wb
• W1 lt W2
• With this approach,
• bs never apply
• as post a bond (W1-Wa) during probation, earn
  a deferred reward (W2-Wa) later
• notice that a-types must trust the firm not to
  cheat them in period 2 more on this later
• it can be shown that the bond reward are larger
• the more accurate the probation sorting q
• the larger the differences in productivity (
  thus market values) between types a b

20
Our Job Offer, Graphically
NPV of job depends on probability of getting
tenure, which differs by employee type
21
More on Signaling

• Our probation model is a special case of
  signaling
• youre talented want to prove it. Invest in a
  signal of your ability
• the signal must be cheaper (easier) for
  high-skilled to obtain than low skilled
• e.g., earning a very difficult degree
• in probation, the signal is willingness to accept
  W1 lt Wa
• examples
• expecting new employees to work very hard to
  prove themselves
• using strong pay for performance
• requiring entrepreneurs to put up capital to get
  venture capital
• requiring partners to invest in joint ventures
• stock buyback by a firm
• Notice that employment is now a long-term
  relationship, rather than a spot market

22
Screening Signaling

• They start w/ different assumptions, but work
  similarly
• in screening, neither knows ability initially
  (imperfect information)
• in signaling, the worker knows (asymmetric
  information), we need to worry about adverse
  selection
• in both, ability is fixed. Well consider
  training soon
• They have similar implications
• workers invest either to find out their ability,
  or to prove it
• probation or similar
• deferred compensation or pay for performance
• Pay often rises more rapidly than productivity
• what role might be played by vesting of pensions
  or other compensation?

23
3. Bidding for Workers

• A specific Director of Marketing candidate has
  been suggested. Should you raid another firm
  for that person?
• there are three possible outcomes

Employee was more valuable at the other firm
You lose
You overpaid Winners Curse
You win
Employee is more valuable at your firm
24
Economics of Raiding

• The cards are stacked against a successful raid.
• If there is a good fit between the employee and
  the firm, it will cost you more to hire her than
  it will for them to keep her.
• Asymmetric information about the prospects true
  ability will yield the winners curse.
• You know less about true ability than the other
  firm does.
• The other firm decides to match your offer or not
• They match when the worker is good. They do not
  match when the worker is not so good.
• You get the workers that are not so good.

25
Possibilities for good raids

• Target workers value is greater in your firm
• You have no one with this skill the other firm
  has several people
• Lee Iacocas move to Chrysler
• Rapid changes in technology mean firms may be at
  different stages
• Target firm is under duress
• Canadian universities
• A life change has occurred for the target.
• Divorce
• Completed degree
• 80 of MBAs switch firms when they receive their
  degree

26
Raiding Other Firms

• When is raiding someone elses specific employee
  more likely to be profitable to you?
• the employees skill mix is unusual, fits your
  circumstances very well
• you happen to know just the right person, saving
  search costs
• something changed recently
• currently employed in declining firm
• industry is evolving rapidly
• just obtained a credential (e.g., MBA)
• divorce
• Otherwise just do regular recruitment (may be
  better to build rather than buy)

27
Matching Outside Offers

• OK, what about the flip side? You have an
  employee who has received an offer. How should
  you respond?
• some firms have a policy of refusing to match any
  outside offers. Should yours?
• If pay is the only relevant factor, it will not
  deter job search
• if he gets an offer you dont match, he quits.
  If you do, hes just as well off
• Two cases where the policy might be worthwhile
• if the employee gets non-pecuniary benefits from
  working for you, refusing to match outside offers
  can deter false job search
• if you overpay the employee, you can only lose
  from his job search
• Risk of not responding might lose your most
  profitable employees
• most firms take a middle ground discourage
  matching outside offers, possibly even stating a
  formal policy, but matching outside offers in
  special cases

28
Who do you worry about losing?

• Employees with external reputations are the most
  likely to receive external offers when they are
  underpaid. You need to know who these employees
  are and you need to have a plan in place in case
  they receive offers.

29
5. Risky Candidates

• Consider two I-Bank job candidates
• salary 100K will work for the firm T years
• Gupta predictably produces 200K per year
• Svensen may be a star (50), producing 500K or
  disaster (50), losing 100K
• ignore discounting for simplicity
• assume the firm is risk neutral (is this a
  reasonable assumption?) where is this more
  likely to be true?
• They have equal expected productivity which is
  the better hire?

30
Risky Hires as Real Options

• The upside potential possibility of firing
  gives the risky hire an option value
• option value NPV (higher career productivity if
  a star)

31
Valuing a Risky Hire

• The option value is higher
• the younger the candidate shorter the trial
  period
• the lower the termination costs
• the larger the upside, smaller the downside
• if you are hiring a portfolio
• the longer you can pay the star less than their
  high productivity
• It may pay to consider hiring a risky candidate
  even if
• has lower expected productivity than safe
  candidate
• the firm is risk averse
• termination costs are high

32
Who Gets the Option Value?

• Can we pay both types the same wage?
• if other firms know you fire poor performers
  keep stars, market value of stars should rise
• risky candidates might even be paid more
  initially they have higher expected value to the
  firm if the firm shares some of the profits they
  create
• Firms profit from risky hires if pay lt
  productivity for stars
• workers have costs of switching jobs
• information about high productivity is not easily
  observed by other firms
• productivity is firm-specific (next lecture)
• large supply of risky candidates
• risky candidates are willing to accept low
  initial wages to signal their ability (see below)
• In practice, managers are often reluctant to hire
  risky candidates. Why?

33
6. Summary

• Key Points that you should consider
• Sorting is more important for high skilled jobs
• look carefully at relevant credentials
• spend more resources on interviews, etc.
• Pre-hire sorting is usually imperfect
• consider some form of probation (possibly
  implicit)
• Consider deferred pay, for performance
• also improves signaling when candidates have a
  good idea of their abilities
• Consider risky hires
• especially when you can capture some of the
  option value diversify risk

34
Economic Ideas

• Employees as real options
• Imperfect information sorting employees
• screening
• Important special case asymmetric information
• adverse selection
• signaling
• moral hazard (later)
• Creating value separately from splitting it
• is there a conflict of interest in the employment
  relationship?