SAP ACADEMIC RESEARCH CONFERENCE€¦ · 3 Time worked routed automatically to project accounting/...
Transcript of SAP ACADEMIC RESEARCH CONFERENCE€¦ · 3 Time worked routed automatically to project accounting/...
SAP ACADEMICRESEARCH CONFERENCEAugust 21, 2009
When Software is Not EnoughSynergies Between Human Resource Practices
and SAP’s HCM Software
Erik Brynjolfsson
MIT Center for Digital Business
SAP Academic Conference Palo Alto, August 21, 2009
IT and Productivity
productivity
Source: Brynjolfsson & Hitt (2003)
IT
3
Why is there so much variation on the return of IT?
IT, HR and ROI• Goal:
– Examine how technology and human resource practicesaffect firm productivity.
• Findings– 3-way system of complementarities is correlated
with higher firm performance1. Human Capital Management (HCM) software2. Monitoring Practices3. Performance Pay Practices
– Any one practice without the others is correlatedwith little or no performance gain
Software is not enough
HCM Software is part of an ERPSystem
• Enterprise Resource Planning (ERP) softwarehas many distinct modules– Firms can choose to adopt the Human Capital
Management (HCM) software module
• HCM is designed to help managers observeand measure performance.– Keep track of time and attendance– Keep track of tasks completed by workers and
time and materials used.– Can generate 90+ reports for management and
employees.5
Example of HCM Reports
6
For each task, the worker recordsa brief description of work and thetime it takes to complete the task.He may also records the materialhe uses to do the task.
Detail Records of Work Management of Time and Attendance
Management can have access todetailed reports to employee’sattendance, such as absence,overtime, illness and vacation.
Human Resource Management
• The Two Key Ingredients of Incentive Systems
1. Performance pay practicesEg. Are compensation plans designed to align pay with
performance and are they linked to KPIs?
2. Monitoring practicesEg. Does the company track the time and attendance of
employees?
7
Complementarities
• Systems of practices may more effective atraising productivity than individual practicealone.– Milgrom & Roberts (1992)– Ichniowski, Prennushi & Shaw (1997)– Brynjolfsson & Milgrom (2009)
• The 3-way combination of HCM, PerformancePay and Monitoring Practices may constitutesuch a system
8
The Archetypal Complementarities Diagram
“1+1=3”
9Practice 1
Pra
ctic
e 2
Lo
Lo
Hi
Hi
0 +
+ +++
Computer Capital is Not Enough
Source: Brynjolfsson, Hitt & Yang (2002)
10
IT
1,1,1
1,0,1
1,1,0
0,1,1
1,0,0
0,0,1
0,1,0
0,0,0,
HCM: F(1,1,1) – F(0,1,1) > F(1,0,0) – F(0,0,0)
X: HCM
Y: Perf PayZ: Monitoring
4 tests of complementaries:
Perf Pay: F(1,1,1) – F(1,0,1) > F(0,1,0) – F(0,0,0)
11
The system:[F(1,1,1) – F(0,1,1)] + [F(1,1,1) – F(1,0,1) ] + [F(1,1,1) – F(1,1,0)] -
[F(1,0,0) – F(0,0,0)] + [F(0,1,0) – F(0,0,0)] + [F(0,0,1) – F(0,0,0)] > 0
Monitoring: F(1,1,1) – F(1,1,0) > F(0,0,1) – F(0,0,0)
The Causality Conundrum:Which Came First, HCM or Productivity?
Technology investment is correlated with performance, but1. Does technology investment cause performance, or2. Does performance lead to more technology investment?
1997
HCM SoftwareGo-live
2002
HCM SoftwarePurchase
The time line of HCM adoption of a firm inthe manufacturing industry for producingmachinery and electronic product.
Fortunately (for researchers, at least!), HCM systemstake a long time to go live. Average = 2.71 years.
Data
• IT Adoption data:– HCM Module Purchase Date
– HCM Module Go live Date
• Human Resource Practices Data– Surveys conducted by ASUG from 2005-2006
– 189 firm responses
– 90 are matched to Compustat
– 15 industry groupings up to 1 ½ SIC.
13
Data—Incentive Systems
• Firm surveys, rate on scale of 1-51. Performance pay (aggregate of 4 questions)
E.g. “Compensation plans are aligned with resource plansto attract and retain the desired skill set.”
2. Performance Monitoring (aggregated 8questions)E.g. “Time and attendance system has automated analysis
and reporting capabilities to analyze KPIs such as losttime, productivity, cost of absence, overtime orillness.”
14
Two Tests of Complementarities1. Correlation Test: The adoption of Incentive Systems components and
HCM should be correlated.
2. Productivity Test: The co-presence of Incentive Systems and HCMshould generate a productivity premium.
*p<.1, **p<.05, ***p<.001
1. Correlation Test
Correlation Tests
16
Dep. Var. HCM HCM HCM
logit All obs. PerfPay>0 PerfPay 0Monitor .460** .175* .500
(.235) (.131) (.765)ControlVariables
IndustryYearSize
IndustryYearSize
IndustryYearSize
Obs. 263 169 45
Dep. Var. HCM HCM HCM
logit All obs. monitor>0 monitor 0Perf Pay -.30* .289* -1.03
(.080) (.150) (.967)ControlVariables
IndustryYearSize
IndustryYearSize
IndustryYearSize
Obs. 461 333 45
*p<.1, **p<.05, ***p<.01, robust standard error
Productivity & Performance
ln(Sales) 1 ln(K) 2 ln(L)
jIndustryControlsjj
kYearkk
4HCMLive 5Monitor 6PerfPay
7(HCMLive* Monitor) 8(HCMLive* PayPerf )
9(PayPerf * Monitor) 10(HCMLive* PayPerf * Monitor)
3HCMPurchase 11(HCMPurchase* Monitor)
12(HCMPurchase* PayPerf )
13(HCMPurchase* PayPerf * Monitor)
Production Function Framework
17
Complementarities: Productivity Test
Dep var: Sales Fixed Effect OLS LinearOLS Fixed Effect1
HCM Install: 0.0315(0.0392)
.001(.068)
-.002(.070)
-.011(.046)
HCM Live: 0.0559*(0.0327)
.115*(.066)
.129*(.068)
.057(.039)
Monitor .090***(.033)
.075**(.0368) --
PerfPay .151***(.046)
.162***(.039) --
HCM Live XMonitor
-.003(.047)
.108***(.032)
HCM Live X perfPay
-.152(.132)
-.129**(.051)
Perf Pay X Monitor .023(.042) --
HCM Live XMonitor X Perf Pay
.266***(.102)
.165**(.069)
Control variables IndustryYear
IndustryYear
IndustryYear
IndustryYear
R2 0.87 .93 .93 .87Observations 396 396 396 384
PRODUCTIVITY: Q = f(K, L, HCM, monitor, perf pay, Live*monitor*perf pay)
*p<.1, **p<.05, ***p<.01, robust standard errors 1 eliminated one outlier
18
Dep Var: lnSales OLS Fixed Effect1
HCM Install: .03(.08)
-.01(.05)
HCM Live: .13*(.07)
.06(.04)
Monitor .07*(.03) --
PerfPay .17***(.05) --
Perf Pay X Monitor .02(.05) --
Install X Monitor -.02(.08)
.06(.07)
Install X perf Pay -.18(.19)
.001(.007)
Install X Monitor XPerf Pay
.17(.18)
-.08(.10)
Live X Monitor -.02(.08)
.12***(.04)
Live X perf Pay .20(.13)
-0.13**(.06)
Live X Monitor Xperf Pay
.27**(.11)
0.14**(0.07)
Control variables IndustryYear
IndustryYear
R2 .93 .87Observations 396 384
Productivity Test: Using HCM Install
Dep var: lnSales OLS Fixed Effect1
HCM Install: -.01(.08)
-.03(.04)
Monitor .09**(.03) --
PerfPay .15***(.05) --
Install X Monitor -.05(.07)
.03***(.06)
Install X perf Pay -.12(.20)
0.10**(.09)
Perf Pay XMonitor
.03(.05) --
Install X Monitor XPerf Pay
.13(.18)
-0.12(0.09)
Control variables IndustryYear
IndustryYear
R2 .93 .87Observations 396 384
*p<.1, **p<.05, ***p<.01, robust standard errors.1 eliminated one outlier
1,1,1
1,0,1
1,1,0
0,1,1
1,0,0
0,0,1
0,1,0
0,0,0,
HCMLive: F(1,1,1) – F(0,1,1) > F(1,0,0) – F(0,0,0)
X: HCM Live
Y: Perf PayZ: Monitoring
4 tests of complementaries:p=.0039
Not RejectedPerf Pay: F(1,1,1) – F(1,0,1) > F(0,1,0) – F(0,0,0)
20
The system:[F(1,1,1) – F(0,1,1)] + [F(1,1,1) – F(1,0,1) ] + [F(1,1,1) – F(1,1,0)] -
[F(1,0,0) – F(0,0,0)] + [F(0,1,0) – F(0,0,0)] + [F(0,0,1) – F(0,0,0)] > 0p=.0038
p=.0041Monitoring: F(1,1,1) – F(1,1,0) > F(0,0,1) – F(0,0,0)
Conclusion1. There are Synergies Between Certain HR
Practices (performance pay and monitoringpractices) and HCM software
1. Firms are more likely to adopt HCM if they alreadyhave the “right” HR Practices in place
“Correlation Test”
2. HCM is correlated with higher productivity whenadopted together with these practices
“Productivity Test”
3. Performance benefits occur when systems “go-live”,not when they are merely purchased.
“Causality Test”21
Implications for Managers
1. The IT “Productivity Paradox” for some firms maybe due to missing complements
2. HCM Installations will be most successful ina) Firms that already have the right HR practices in place,
and
b) Firms that are willing to make organization co-investments
3. Over time, technology and organizational practicesmust co-evolve
The Bottom Line: Software is not enough!22
To learn more:http://digital.mit.edu/erik
and
Aral, Brynjolfsson and Wu “Testing Three-WayComplementarities: Performance Pay,Monitoring and Information” WorkingPaper, MIT Center for Digital Business,
July, 2009.
24
When monitoring is high, monitoringand HCM software are complement
When monitoring is low, monitoring andHCM software are complement
3D Graphs of Complementarities
Incentive Compensation Practices
1 Compensation plans are designed to support overall corporatebusiness strategy as well as strategies of individualdivisions/departments.
2 Compensation plans are designed to align pay withperformance, and are linked to easily understood KPIs (e.g.,corporate, divisional, organizational profitability).
3 Compensation plans are aligned with resource plans to attractand retain the desired skill set.
4 Employee performance expectations clearly communicatedduring Recruiting process.
5 Compensation system provides comprehensive support fordisplaying and editing employee compensation-related data,activating planned and approved compensation adjustments,and transmitting the information to payroll.
Performance Monitoring Practices I
1Compensation planning system integrates information with otherrelevant non HR systems, such as financial systems, OSHA,manufacturing, sales.
2 Standardized job descriptions and job evaluations are availableonline.
3 Time worked routed automatically to project accounting/ resourceplanning systems.
4Time and attendance system has automated analysis andreporting capabilities to analyze KPIs such as lost time,productivity, cost of absence, overtime or illness.
Performance Monitoring Practices II
6HR system allows for a Balanced Scorecard framework which isintegrated into department and individual performance appraisaldocuments and supports benchmarking and continuousimprovement.
7 HR System provides data analysis and reporting tools to supportHR policy development and decision making.
8HR system allows managers to analyze workforce data; todesign, implement and monitor corporate strategies to optimizethe workforce; and to continuously evaluate how various coursesof action might affect business outcomes.
9HR system enables HR professionals to develop cost effectiveresource strategies, by supporting accurate planning process,allowing monitoring of actual performance relative to plan andallowing simulations of multiple planning scenarios or analysis.
SAP ACADEMICRESEARCH CONFERENCEAugust 21, 2009