Why Is Change at Work So Hard?

The old system had an export that broke every quarter and a workaround everyone complained about. The replacement fixes both, and three weeks after launch half the team still uses the old way.

Change at work is hard for three reasons. The brain's defenses fire early (Nesse, 2005) and respond to the unfamiliar (Cunningham and Brosch, 2012). A change at work also threatens standing and control, the two strongest measured stress triggers (Dickerson and Kemeny, 2004). And leaving a current arrangement costs the brain more than keeping it (Fleming, Thomas and Dolan, 2010).

What Makes Something Feel Safe to the Brain?

Repetition alone changes how a person feels about a thing. Robert Zajonc showed in 1968 that repeated exposure to unfamiliar stimuli, from invented words to characters in an unknown script to faces, raised how much people liked them, and that liking rose with the number of exposures, an effect now called the mere exposure effect (Zajonc, 1968). Robert Bornstein then pooled 208 experiments from the two decades that followed and found the effect reliable across stimulus types, with an overall effect size of .26 (Bornstein, 1989). The effect was larger when people were unaware of the exposures, recognizing a stimulus as one you had seen before tended to weaken it, and it peaked at roughly 10 to 20 presentations.

So the old process has an advantage that has nothing to do with its quality. It has been encountered far past the point where the exposure effect peaks, and the new process has been encountered once, at the announcement. The laboratory figure comes from ratings of simple stimuli, and the meta-analysis covers no workplace stimuli, so the number gives the direction of the effect and its rough scale.

Why Does the Unfamiliar Register as a Possible Threat?

Randolph Nesse applied signal-detection theory to the body's defenses and showed that when a false alarm costs little and a missed threat costs a great deal, the threshold that minimizes total cost is low, so the defense fires often when nothing is wrong (Nesse, 2005). He named this the smoke detector principle. Applied to the unfamiliar, it predicts a response to a new situation before there is evidence about it.

Imaging data fit that prediction. The amygdala, the structure behind the detection step that Why Do I React to an Email Before I've Finished Reading It? describes, responds to novel and ambiguous events as well as threatening ones (Cunningham and Brosch, 2012). In the first week after a launch, every screen is new and ambiguous at the same time, so each one draws the amygdala's response.

Why Does the Current Arrangement Win?

The economists William Samuelson and Richard Zeckhauser gave people hypothetical choices in two versions, one with the options presented neutrally and one with an option labeled as the current state. The labeled option drew more choices, and its relative advantage grew as more alternatives were added to the choice set (Samuelson and Zeckhauser, 1988). Their field data showed the same pattern. In Harvard's health plan records, employees who had been enrolled for years kept the incumbent plan far more often than new enrollees choosing from the same menu. They named the pattern status quo bias.

Leaving a default also costs the brain measurable work. Stephen Fleming, Charlotte Thomas and Raymond Dolan scanned 16 people making difficult judgments about whether a tennis ball had landed in or out, with one answer set as the default on each trial. As the judgments got harder, people accepted the default more often, and on the hard trials errors were more frequent when the default was accepted. Rejecting the default on a hard trial raised activity in the subthalamic nucleus, a structure involved in stopping and switching actions (Fleming, Thomas and Dolan, 2010). The study varied the difficulty of a perceptual judgment and did not test workload or fatigue. Stress does change how people decide, in ways that depend on the task (Starcke and Brand, 2012), as Why Do I Make Worse Decisions When I'm Stressed? describes.

Loss weighting adds to the pull. Daniel Kahneman and Amos Tversky found that in choices under risk "losses loom larger than gains" (Kahneman and Tversky, 1979). A change presents what will be given up alongside what will be gained, and the loss is weighted more heavily.

Why Does a Change Announced without Details Raise Anxiety?

A change announced without detail leaves people with an outcome they cannot predict, and the research separates that condition from a known bad outcome. In Christian Grillon's laboratory work, an unpleasant stimulus that could arrive at any moment raised background anxiety, measured by the startle reflex, above the level produced by the same stimulus announced by a cue (Grillon, Baas, Lissek, Smith and Milstein, 2004). In a separate EEG study, participants high in neuroticism showed a larger brain response to uncertain feedback, a question mark, than to clearly negative feedback (Hirsh and Inzlicht, 2008). That pattern was specific to those participants, so the discomfort of not knowing varies from person to person.

The laboratory stress research adds a second condition. Across 208 studies, the stressors that produced the largest cortisol responses, and the slowest recovery, combined being judged by others with having no control over the outcome (Dickerson and Kemeny, 2004). An announced change with the details withheld supplies both conditions at once, uncertainty about the outcome and a judgment that nothing the person does in the meantime affects.

What Does a Workplace Change Threaten besides the Process?

The new system is rarely the whole threat. Dickerson and Kemeny explain their finding with social self-preservation theory, set out in What Does Fight or Flight Mean at Work?, in which standing among other people is defended by the same stress systems that respond to physical danger (Dickerson and Kemeny, 2004). A new system asks people who are competent on the old one to become beginners in front of their colleagues, which puts standing at risk, and it was decided elsewhere, which removes control. Exclusion has a measured cost of its own, in a brain region tied to the distress of pain (Eisenberger, Lieberman and Williams, 2003), the finding behind why being left out of a meeting hurts.

David Rock's SCARF model, a practitioner model, lists status, certainty, autonomy, relatedness, and fairness as the social conditions people at work defend (Rock, 2008). Rock published it without original data, and the peer-reviewed evidence sits behind the separate conditions, in Dickerson and Kemeny's review and in the exclusion study above. A reorganization can change all five conditions in one announcement.

Why Does the Change Still Feel Unsettled after It Has Happened?

William Bridges, a practitioner, separates the change, the event with a date, from the transition, the adjustment inside each person, which in his account continues after the launch and has no fixed length (Bridges, 2009).

Whether people can say the new way is not working depends on the team around them, a condition Amy Edmondson measured as psychological safety in 51 work teams (Edmondson, 1999), and that condition is part of why some teams speak up and others stay silent. In a team where the first person to say "I cannot find the report" is answered with help, the problem gets reported.

A sequence of changes has a cost of its own. Change fatigue, the perception that too much change is taking place, has a validated measure and is associated with exhaustion and lower commitment (Bernerth, Walker and Harris, 2011), the condition a team describes when it is exhausted by another change.

Is the Brain Set against Change?

No. The same person who resists a new reporting tool will try an unfamiliar restaurant that evening, and the features of the two situations account for the difference. The restaurant is chosen, reversible, low in consequence, and nobody is grading the choice. The reporting tool is imposed, hard to undo, carries consequences, and is used in front of colleagues, which supplies the two conditions from Dickerson and Kemeny's review above. The amygdala's response to novelty runs for positive events as well as threatening ones (Cunningham and Brosch, 2012), so novelty by itself does not produce resistance.

People do differ in how strongly they react to imposed change. Shaul Oreg developed a measure of dispositional resistance to change across seven studies and found four factors, which he named routine seeking, emotional reaction to imposed change, short-term focus, and cognitive rigidity (Oreg, 2003). The factors describe tendencies on a scale, so the same change produces a range of reactions in one team. Behind a refusal, the research above points to a specific loss, of standing, control, predictability, or competence on the old process, and a refusal treated as a response to that loss gives the person leading the change something to act on.

How Does the Brain Learn That the New Way Is Safe?

The same mechanism that favors the old way builds comfort with the new one. Each encounter with the new process is one more exposure toward the peak Bornstein measured.

Practice has its own timeline. Phillippa Lally and colleagues followed 96 volunteers who each chose a daily behavior to make habitual and tracked how automatic it became over 12 weeks. Among the 39 participants whose data fit the curve, the median time to reach the automaticity plateau was 66 days, with a range from 18 to 254, and a single missed day did not materially set the process back (Lally, van Jaarsveld, Potts and Wardle, 2010). The brain changes physically during that practice. In a study of 24 people, the 12 who learned to juggle for three months showed expanded gray matter in a visual motion area and in the parietal cortex, and the expansion partly receded after three months without practice (Draganski and colleagues, 2004). In that study the adult brain changed with practice and the change partly reversed without it, and the realistic timeline for a new habit is weeks to months.

How Do the Three Causes Play Out over a Change?

The three causes arrive in a sequence, and each phase of a change has one of them in front.

At the announcement, the uncertainty research applies. People learn that something will change, often before they learn what it means for them, which is the unpredictable condition Grillon measured, combined with a judgment they cannot affect. The body response starts with the announcement itself, and what happens in your body when someone says "we need to talk" is the same sequence at the scale of one conversation.

At the launch, novelty and the default cost apply. Every screen is new, so the detection step runs at each one, and every task now requires rejecting a familiar default, which in Fleming's data was hardest to leave when the judgment was hardest. In the scenario above, this is the week in which the old export is most likely to be used.

In the weeks after, the exposure and practice research applies. Each use is an exposure and a practice repetition, so the novelty response fades and automaticity builds over the weeks Lally measured, with the transition continuing alongside. The three causes do not leave in order. Standing and control remain active for as long as people are using the new system in front of colleagues, so the second cause runs through all three phases.

What Does This Mean for Anyone Leading a Change?

Each of the three causes points to a move, and the transition and recovery research add two more.

Build familiarity before the launch. Demonstrations, pilots, screenshots, and a colleague using the new system in a regular meeting are exposures, and an announcement is one.

Reduce the two conditions that drive the largest response. What people know about their standing and what they can still decide are the two things an announcement can change, and the wording of an announcement sets both, which is one reason the same message calms one team and panics another.

Lower the cost of leaving the default. Schedule the switch away from the heaviest week of the quarter, and break it into steps, since rolling out change in small steps works better when each step is a smaller decision.

Plan for the weeks after the launch. Put them on the project plan, with a date to collect what is not working and a named person whose job is to answer the first "I cannot find the report."

Protect recovery. Bruce McEwen named the wear from repeated or poorly regulated stress responses allostatic load (McEwen, 1998), and a sequence of changes with no gap between them is one way to keep the response running for a long time. At the individual scale, that is the point when stress stops helping and starts wearing you down. Recovery between demands is one of the practices behind how you stay calm under pressure at work, and for a team it means a gap between one change and the next.

What Status Quo Bias Means

Status quo bias: the tendency to keep the existing arrangement more often than its qualities alone would predict. Samuelson and Zeckhauser demonstrated it in 1988 with labeled options and with health plan records (Samuelson and Zeckhauser, 1988). At work it appears as a pull toward the old process after the new one has proved better.

What Does the Popular Version Get Wrong?

The popular account says people are hardwired to resist change. The evidence shows a response to specific features of a situation. The amygdala's novelty response is not specific to threat (Cunningham and Brosch, 2012), liking changes with exposure and no new information (Bornstein, 1989), and the disposition to resist imposed change varies along a scale (Oreg, 2003). The popular account also treats the resistance as fixed. Lally's habit research and Draganski's imaging show habits becoming automatic and the brain changing physically over weeks of practice (Lally and colleagues, 2010; Draganski and colleagues, 2004). And the popular account often arrives with a reptilian-brain story that the anatomy does not support (Cesario, Johnson and Eisthen, 2020), one of the patterns in how you spot bad brain science.

For the next change you are involved in, count how many times the people affected have seen the new way before they are expected to use it, and write down what the announcement told them about their standing and about the decisions left to them.

FAQ

Why do people resist change even when it is an improvement?

Because the old way carries an advantage that has nothing to do with quality. Repeated exposure raises liking with no new information, an effect found reliable across 208 experiments (Bornstein, 1989), and an option labeled as the current state draws more choices than the same option presented neutrally (Samuelson and Zeckhauser, 1988). The new way also carries an outcome people cannot yet predict, and unpredictable threat raised anxiety more than predictable threat did in the laboratory (Grillon and colleagues, 2004).

Why does a well-explained change still get resisted?

Because an explanation changes what people know, and the measured causes act elsewhere. Liking rises with exposure count and no new information (Bornstein, 1989). Leaving a default costs effort that rises with the difficulty of the decision (Fleming, Thomas and Dolan, 2010). And being judged with no control over the outcome drives the largest stress response (Dickerson and Kemeny, 2004), so an explanation that gives reasons without addressing standing and control leaves the two strongest stress triggers in place.

Is resisting change a personality problem?

No. The mechanisms appear in experiments with volunteers who were facing no workplace change, who preferred the familiar (Bornstein, 1989), found uncertainty uncomfortable (Grillon and colleagues, 2004), and accepted a default when a decision was hard (Fleming, Thomas and Dolan, 2010). People do differ in degree. Oreg's measure of dispositional resistance to change has four factors, routine seeking, emotional reaction to imposed change, short-term focus, and cognitive rigidity, and they describe tendencies on a scale (Oreg, 2003).

Why is uncertainty about a change so uncomfortable?

Unpredictable threat raises background anxiety more than predictable threat. In a study of 72 people, anxiety between cues, measured by the startle reflex, rose from no threat to predictable threat to unpredictable threat (Grillon and colleagues, 2004). A change announced without a timeline or a decision is the unpredictable condition, and nothing the person does in the meantime changes the outcome.

Does the brain work harder to change than to stay the same?

In one setting, yes. When people made difficult perceptual judgments with a default answer available, rejecting the default raised activity in the subthalamic nucleus, and people accepted the default more often as the judgments got harder, including when the default was wrong (Fleming, Thomas and Dolan, 2010). Leaving the default carries a cost that rises with the difficulty of the decision.

Why does the same person try a new restaurant but resist a new system at work?

Because the two situations differ in the features that set off the stress response. The restaurant is chosen, reversible, and ungraded. The system is imposed, consequential, and used in front of colleagues, which supplies the two strongest laboratory triggers, being judged and having no control (Dickerson and Kemeny, 2004). Novelty alone does not produce resistance; the amygdala responds to positive novelty too (Cunningham and Brosch, 2012).

How long does it take for a new way of working to feel normal?

Weeks to months. In a field study that modeled it, the median time to the automaticity plateau for a new daily behavior was 66 days, with a range of 18 to 254 among the 39 participants whose data fit the curve (Lally and colleagues, 2010). The exposure effect on liking peaks after roughly 10 to 20 encounters (Bornstein, 1989).

What can a person leading a change do about this?

Three things that match the three causes. Add exposures before the launch, since familiarity is built by repeated contact (Bornstein, 1989). Say what is changing, what is staying the same, and what people can still decide, because being judged and having no control are the strongest triggers (Dickerson and Kemeny, 2004). And time the switch away from the busiest week, because leaving a default costs most when the decision is hardest (Fleming, Thomas and Dolan, 2010).

Related Posts:

What Does Fight or Flight Mean at Work?

Why Do I React to an Email Before I've Finished Reading It?

What Happens in Your Body When Someone Says We Need to Talk?