Monday, July 13, 2015

Fixing General Motors’ Culture—Any Lessons for Nuclear?

GM Headquarters
In a recent interview* with LinkedIn, General Motors CEO Mary Barra discussed her plan for fixing GM’s culture.  The interviewer asked what needs to change, what about known problems like the “GM nod”** and the siloed organization, and what is the key to the improvement process?  The following quotes are excerpted from her answers.  Do they suggest a clear vision for the future culture and/or a satisfactory action plan?

“. . . get everyone engaged, working together, and bringing the best ideas forward[.]”

“. . . I never accepted the GM nod.  If somebody said in a meeting they were going to do something, I expect you to do it.”

“We've got to model [working across the organization].”

“[We have to] own each other's problems.”

“So our goal is to be the safety leader. We're really driving a zero-defect mentality.”

“If we can get in a room and really, you know, argue it out constructively and everybody's views get on the table, we'll make better decisions.”

“. . . we've got to earn the trust of every single employee by demonstrating the way we behave.”

Our Perspective

We realize this was not some carefully crafted article for the Harvard Business Review but there are too many soft spots in this recipe for fixing the culture to let this interview slide by without comment.

Let’s begin with the positives.  Barra promotes respect for ideas; that’s a positive feedback loop and a good thing.  Senior management modeling desired behavior and working to earn employee trust are both essential for cultural change.  Safety leadership is certainly a laudable goal.  

The nod is a little more problematic.  Maybe Barra never accepted the nod but plenty of other folks did.  Is modeling the desired behavior sufficient to create change?  How long will it take?  What else might need to be done?

Shared ownership of problems is a good start but how does GM establish, model and inculcate a process that obtains permanent problem resolutions going forward?

Barra also believes an insider (like her) is better suited for changing the culture than an outsider.  We agree an insider may have a better handle on recognizing when employees are trying to spin a situation in their favor but an outsider can bring a clear view of the performance gap between an organization’s current state (e.g., its characteristics, priorities and processes) and where it needs to be.

Some ingredients are missing.  Most importantly, there is no mention of the powerful cost/finance feedback loop that contributed to GM’s quality problems.  Wringing pennies out of product costs was a major goal for years.  What roles will cost consciousness and management financial incentives play going forward?

In another area, how is the management decision making process changing other than arguing things out?

Bottom line: There are no lessons for nuclear in the GM CEO’s outline of her cultural change initiative.  In fact, her proclamations sound just like nuclear managers’ braying when they try to convince regulators, the media and the public that something, anything is happening to address perceived cultural issues.  But what usually isn’t happening is some in-depth analysis of how their organizational system functions.

*  D. Roth, “Mary Barra's Got a Plan for Fixing GM's Culture (and Only an Insider Can Pull it Off),” LinkedIn interview (July 6, 2015).  Safetymatters co-founder Bob Cudlin first spotted and called attention to this article.

**  The “GM nod” was “where employees would commit to being on board with a decision, then ignore it [later.]”

Friday, July 3, 2015

New Safety Culture Assessment at the Hanford Waste Treatment Plant

Hanford WTP
The Department of Energy (DOE) recently released the latest safety culture (SC) assessment report* for the Hanford Waste Treatment Plant (WTP or “vit plant”) project.  The 2015 report follows similar SC assessments conducted in 2011 and 2014, all of which were inspired by the Defense Nuclear Facilities Safety Board’s scathing 2011 report on SC at the WTP.  This post provides a brief overview of the report’s findings then focuses on the critical success factors for a healthy SC.

Assessment Overview

The 2011, 2014 and 2015 assessments used the same methodology, with multiple data collection methods, including interviews, Behavioral Anchored Rating Scales (BARS)** and a SC survey.  Following are selected highlights from the 2015 report.

DOE’s Office of River Protection (ORP) has management responsibility for the WTP project.  In general, ORP personnel feel more positive about the organization’s SC than they did during the 2014 assessment.  Feelings of confusion about ORP’s more collaborative relationship with Bechtel (the prime contractor) have lessened.  ORP management is perceived to be more open to constructive criticism.  Concerns remain with lack of transparency, trust issues and the effectiveness of the problem resolution process.

Bechtel personnel were more positive than in either previous SC assessment.  Bechtel has undertaken many SC-related initiatives including the promotion of a shared mental model of the project by senior Bechtel managers.  In 2014, Bechtel Corporate’s role in project decision making was perceived to skew against SC concerns.  The creation of a new Bechtel nuclear business unit has highlighted the special needs of nuclear work. (pp. 2, 39)  On the negative side, craft workers remain somewhat suspicious and wary of soft retributions, e.g., being blamed for their own industrial mishaps or having their promotion or layoff chances affected by reporting safety issues.

See this newspaper article*** for additional details on the report’s findings. 

Critical Success Factors for a Healthy SC

We always look at the following areas for evidence of SC strength or weakness: management’s decision making process, recognition and handling of goal conflicts, the corrective action program and financial incentives.

Decision Making

Both ORP and Bechtel interviewees complained of a lack of basis or rationale for different types of decisions. (pp. 9, 16)  Some ORP and Bechtel interviewees did note that efforts to clarify decision making are in process. (pp. 13, 32)  Although the need to explain the basis for decisions was recognized, there was no discussion of the decision making process itself.  This is especially disappointing because decision making is one of the possible behaviors that can be included in a BARS analysis, but was not chosen for this assessment.

Goal Conflicts

Conflicts among cost, schedule and safety goals did not rise to the level of a reportable problem.  ORP interviewees reported that cost and schedule do not conflict with safety in their individual work. (p. 6)  Most Bechtel interviewees do not perceive schedule pressures to be the determining factor while completing various tasks. (p. 23)  Overall, this is satisfactory performance.

Corrective Action Program

We believe how well an organization recognizes and permanently resolves its problems is important.  Problem Identification and Resolution was one of the traits evaluated in the assessment.  ORP interviewees said that current safety concerns are being addressed.  The historical lack of management feedback on problem resolution is still a disincentive for reporting problems. (pp. 8-9)  Some Bechtel interviewees said “issue resolution with management engagement was the single most positive improvement in problem resolution, . . .” (p. 24)  This performance is minimally acceptable but needs ongoing attention.

Financial Incentives

DOE’s contract with Bechtel now includes incentives for Bechtel if it self-identifies problems (rather than waiting for DOE or some other party to identify them).  ORP believes the incentives are a positive influence on contractor performance. (p. 8)  Bechtel interviewees also believe the new contract has had a positive impact on the project.  However, Bechtel has a goal to reduce legacy issues and some believe the contract’s emphasis on new issues distracts from addressing legacy problems. (pp. 24-25)  The assessment had no discussion of either ORP or Bechtel senior management financial incentives.  The new contract conditions are good; ignoring senior management incentives is unacceptable.

Safety Conscious Work Environment (SCWE)

We usually don’t pay much attention to SCWE at nuclear power plants because it is part of the larger cultural milieu.  But SCWE has been a long-standing issue at various DOE facilities, as well as the impetus for the series of WTP SC assessments, so we’ll look at a few highlights from the SC survey data.

For ORP, mean responses to five of the six SCWE questions were higher (better) in 2015 vs 2014, and 2014 vs 2011.  However, for one question “Concerns raised are addressed” the mean is lower (worse) in 2015 vs 2014, and significantly lower in 2015 vs 2011.  This may indicate an issue with problem resolution. (p. B-2) 

For Bechtel, mean responses to all six SCWE questions were significantly higher (better) in 2015 vs 2014.  However, the 2011 data were not included so we cannot make any inference about possible longer-term trends. (p. B-5)  What is shown is good news because it appears people feel freer to raise safety concerns.  Interestingly, Bechtel’s mean 2015 responses were 5-13% higher (better) than ORP’s for all questions.

Both ORP and Bechtel are showing acceptable performance but continued improvement efforts are warranted.

Our Perspective

The Executive Summary and Conclusions suggest ORP and especially Bechtel have turned the corner since 2014. (pp. v, 37)  This is arguably true for SCWE but we’d say the jury is still out on improvement in the broader SC, based on our look at the BARS data.

For ORP, the BARS data mean scores are higher for 4 (out of 10) behaviors in 2015 vs 2014, but only higher for 1 behavior in 2015 vs 2011. (p. B-1)  The least charitable interpretation is ORP’s view of itself has not yet re-achieved 2011 levels.  For Bechtel the BARS data shows a bit brighter picture.  Mean scores are higher for 6 (out of 10) behaviors in 2015 vs 2014, and higher for 4 behaviors for 2015 vs 2011. (p. B-4)

The format of the report is probably intended to be reader-friendly but it mixes qualitative interview data and selected quantitative data from BARS and the survey.  The use of modifiers like “many” and “some” creates a sense of relative frequency or importance but no real specificity.  It’s impossible to say how much (if any) cherry picking of the interview data occurred.****

We also wonder about the evaluation team’s level of independence and optimism.  This is the first time DOE has performed a WTP SC assessment without the extensive use of outside consultants.  Put bluntly, how independent was the team’s effort given DOE Headquarters’ desire to see improvements at WTP?  And it’s not just HQ; DOE is under the gun from Congress, the DNFSB, the Government Accountability Office, and environmental activists and regulators to clean up their act at Hanford.

We want to see a stronger SC at Hanford but we’ll go with Ronald Reagan on this report: “Trust, but verify.”

*  DOE Office of Enterprise Assessments, “Follow-up Assessment of Safety Culture at the Hanford Site Waste Treatment and Immobilization Plant” (June, 2015).  We have followed the WTP saga for years; please click on the Vit Plant label to see our related posts.

**  Behavioral Anchored Rating Scales (BARS) quantitatively summarize interviewees’ perceptions of their organization using specific examples of good, moderate, and poor performance.   There are 17 possible organizational behaviors in a BARS analysis, but only 10 were used in this assessment:  Attention to Safety, Coordination of Work, Formalization, Interdepartmental Communication, Organizational Learning, Performance Quality, Problem Identification and Resolution, Resource Allocation, Roles and Responsibilities and Time Urgency. (p. C-2)

***  A. Cary, “DOE: Hanford vit plant safety culture shows improvement,” Tri-City Herald (June 26, 2015).

****  The report also includes multiple references to the two organizations’ behavioral norms that were inferred from the survey data.  It’s not exactly consultant mumbo-jumbo but it’s too complicated to attempt to explain in this space.