Since the Roman Empire: Organizations Engage History to Make Changes

How to change an organization? The answer to this question is made surprisingly difficult by all those who think that change is unnecessary, change is risky, and in any case it should be change that favors their favorite option; nothing else will do. Using history to promote a change effort is an old trick that makes a lot of sense, because it is a way of claiming that change is actually a return to a golden age. And history can be edited in many ways, so it is a very flexible trick. Managers use it.
But can it be more than a managerial trick to manipulate the organization? New research by Mary Jo Hatch and Majken Schultz in Administrative Science Quarterly shows how change can be created in a more autonomous fashion by employees reaching back into the organizational history. The research follows two distinct and independent occasions that Carlsberg brewery used its old motto, the latin phrase semper ardens (always burning) to foster change. In each case the users were different and the change was different, but the old and flexible motto proved a way to successfully make changes with less controversy. In one case, a group of master brewers working on their own used it to formulate, gain acceptance for, and launch a craft beer line, in stark contrast to the industrial beer that was the core of Carlsberg. In the other, it was proposed by consultants seeking to create a unifying statement for Carlsberg, which had become large and diverse through recent mergers,and then promoted internally in the organization.
Even though these processes were unrelated, even to the extent the consultants were unaware of the earlier semper ardens use, they followed a remarkably similar sequence.  The steps are described in detail in the paper, but here I want to focus on the two final ones: renewing and re-embedding. Renewing is central when history is used to motivate change, because the new activities are never exact equivalents of the historical record. Indeed, the historical record can be unclear or even contradictory, so renewal is needed. Semper ardens was a phrase favored by the second generation Carlsberg owner, but did not have any concrete brewing practices associated with it. But the master brewer team reached back into the brewing recipes from that time period, and combined these with the passion for improvement expressed through the “always burning” meaning to create beers that were distinct in taste and packaging.  
Re-embedding is actions taken to give the referral to history endurance in the organization. This is needed because the change attempts are frequent and often override previous ones, including those backed by history, so without embedding changes may become temporary. The master brewer team were able to embed semper ardens into the organization well enough that it lived on in a new craft brewery project even after the beer using it as a label was discontinued, and as a marker of distinction used when announcing extraordinary team efforts or noteworthy events. Thus the motto lived on in its renewed form of encouraging a passion for improvement at Carlsberg. And passion for improvement is, we might agree, useful both for organizations in general and for beer brewers specifically.

Hatch, Mary Jo and Majken Schultz. 2017. Toward a Theory of Using History Authentically: Historicizing in the Carlsberg Group. Administrative Science Quarterly, forthcoming. 

Fertilizing Green Chemistry: How an Occupation Renews Itself

Green chemistry is a set of principles to make chemistry healthier, safer, and more environmental. It has made significant changes in how chemistry is done, and in a way that is very different from how we usually think of reforms in organizations. Usually we think of a powerful outside actor starting reform, like the state, not individuals in an occupation. Usually we think of reforms as a series of prescribed practices, not as principles that each actor translates into practices. With no powerful actor to prescribe and no practice to prescribe, it seems like a puzzle that green chemistry could even become important. How did it happen?
The answer is connected to a set of mechanisms that describe how occupations can change themselves, and is described in a paper by Howard-Grenville, Nelson, Earle, Haack, and Young in Administrative Science Quarterly. The method behind the mechanisms is based on two ideas: change is voluntary, and the people in the occupations are highly diverse. As a result, many people with different views and professional practices need to be persuaded. This is a common problem for social movements that seek to reform occupations from inside, so it is broader than the specific case of green chemistry. And, it is a complicated problem too, as green chemistry showed.
The advocates of green chemistry used three methods: 1) portraying it as normal, 2) explaining how it was morally right, and 3) saying it was a pragmatic approach. Each of these three methods had limited success because they were tailored to specific chemist roles, but in total they were effective because they covered key roles in the occupation. Normal portrayals matched the chemistry professionals as being innovators. Moralizing worked through the teaching role that many of them had. Pragmatic matches their role is industrial problem solvers.  Any chemistry professional might spend some time in each of these roles, and some were heavily dedicated to one of them, making them open to influence through these different methods.
But using three methods has one disadvantage: they are inconsistent, and this is easy to recognize. The inconsistency of the methods also led to inconsistency in the principles – imagine how much stricter the moral approach was – and lack of clarity in how one could do green chemistry. However, the inconsistency was not enough to make green chemistry fail. Because many chemistry professionals accepted these principles, as a result of any of the three methods of persuasion, they instead turned their focus on how to make them consistent – either by finding ways to integrate them, or by finding ways of switching focus depending on circumstances.

This is interesting because it suggests that internal reform does not work the same way as a political movement. Political movements thrive on apparent consistency in messaging and principles, and will typically try to solve (or deny) inconsistency before turning to advocacy. Occupational movements do not need consistent messages, but rather that each individual occupation member is convinced. They will later find ways to solve the inconsistency, because occupational members – unlike political movements – encounter inconsistency in their daily work and are used to finding solutions. Green chemistry has succeeded through a combination of persuasion methods followed by problem solving.

War, Exploration, and Interference: The Rise of Amateur Broadcasters

In daily life we know that professionals rule the roost. Anything remotely important is done by a profession with restricted access to practice and many rules for practitioners — or it is done illegitimately. Did you undergo medical treatment last time you were ill, or did you see a homeopath? Many activities that seem easier and safer also take on profession-like features. Espresso making is done by a high-pressure machine, but there is still a barista profession with formal training and certification. Researchers also have been interested in professions, especially because their effects range from regulating the safety and quality of important service (again, think doctors) to restricting access to work in a way that looks like a power grab (pick your favorite example).
So is there room for non-professionals to get things done? Gregoire Croidieu and Phillip Kim answer that question in a recent article in Administrative Science Quarterly, looking at the key role of amateurs in the development of radio broadcasting in the US. They show that amateurs can get a significant role if the right conditions are in place, even as professionals, companies, and the state seek to push them to the margins. How? Well, that’s where the war, polar exploration, and interference come in.
Let’s start with interference. Technically that is what happens when radio transmitters are near each other in signal spectrum and physical space, and distort each other’s transmissions. It was a major reason that many sought to limit access to the airwaves of amateurs, especially those building their own transmitters and behaving independently from the profession. Socially the limitation of access was also a form of interference – trying to make it hard to be an amateur. But radio amateurs were enthusiastically building up their lay expertise and using it, legally or not. Except for the WWI years, they could be given access as registered radio operators.
That brings us to the war. WWI was when radio amateurs were blocked from the airwaves, with security given as the reason, but it did not mean that they stopped broadcasting. They signed up for military service instead, and fully half of the military radio operators were originally amateurs. This was when the state recognized the value of the lay experts, and took advantage of their skills. After the war, they were supposed to return to their old status as marginal actors, more than before (rising to 20,000 in 1922), but still regulated and limited. Professional radio operators still campaigned against amateurs, seeing them as having little value.
This is where the polar explorations come into play. The amateurs were many, highly skilled, and willing to experiment, and they soon registered a series of technical accomplishments – including shortwave communications with the North Pole, which had been thought impossible. The amateurs, through their lay expertise, became leaders in radio. This role soon turned into the start of radio as an industry and as lay culture, because the establishment of radio stations for communicating to many – instead of point-to-point – happened in parallel. Radio ownership and interest in radio listening rose also, and the radio broadcasting industry eventually grew to as many radio stations as there were licensed radio operators in 1921.
War, exploration, and interference were three of the elements that brought amateurs to the forefront of radio, against the resistance of professionals, companies, and the state. Clearly it was not an easy process, and it took a lot of interest to gather the necessary momentum. Does this show that amateurs have a clear role in society, or that they can overcome the odds under special circumstances? We clearly need to learn more about this so we can understand when activities become professionalized, and when they are open to amateurs.


Croidieu, G., Kim, P.H. (2017). Labor of Love: Amateurs and Lay-expertise Legitimation in the Early U.S. Radio Field. Administrative Science Quarterly

Albert Dunlap Style Likability: Those Who Seek Flattery Get Enemies

I will start this post with an old story. CEO of Sunbeam Corp., Albert Dunlap, known as an expert in turning around troubled firms and selling them for a profit, was sued by the SEC in 2001 for accounting fraud. He was eventually barred from serving as an officer or director in any company, plus ordered to pay investors defrauded money in a class-action lawsuit.  Albert Dunlap was clearly someone in need of flattery, not just money, as he had the classical flattery-sickness symptom of a book written to celebrate his successes (see also his picture!). How he managed things internally in each firm he led is disputed, but much was said about his intimidation of other managers, who probably would conclude that a lot of flattery and ingratiation might help their career. Of course, managers still did better than employees, because his signature move in turning firms around was mass layoffs.
An interesting detail of his downfall was that managers around him were quick to release information that helped the investigation, which is distinct from the many firms with management teams that do all they can to deter and obstruct investigators. Is there a systematic reason for this difference? Possibly. A recent article in Administrative Science Quarterly by Gareth Keeves, James Westphal, and Michael McDonald looks at what happens when managers ingratiate their CEO through flattery and other tools. Their findings are interesting. First, managers who flatter lose their liking of the CEO. Somehow when people artificially put others on a pedestal they also start looking down on them.
Second, managers who flatter may go on to undermine the CEO. The light-handed version of this is to undermine the CEO’s messages to journalists, as this research showed. The heavy-handed version is what happened to Albert Dunlap. Among other events, his comptroller reported that he had been pushing for accounting practices that crossed the legal boundary, and sales people were quick to report “channel stuffing.” Channel stuffing is to sell too many goods and selling them too early, which is not illegal in itself (the sales channel can return unsold goods, so it is safe for them), but it is illegal when the sales are accounted as if they were final.  Those were practices that the SEC (and some investors) suspected, and that meant that what looked like a turnaround in sales and profits was actually a fraudulent scheme.
Seeking flattery is never thought of as a good thing. What we now know is that it also triggers undermining, and for those who have real weaknesses – like a CEO engaged in fraud – that undermining can be very consequential.

Who Admires Martin Shkreli? How we respond to Controversies about Firms

Martin Shkreli is listed as an entrepreneur, hedge fund founder, and pharmaceutical executive. All of these are good things, at least to people who appreciate the formation of new ventures, financing of ventures, and work to improve healthcare. He has also been described as the “most hated man in America.” The trigger for this was when his company Turing Pharmaceuticals acquired the drug Daraprim, an essential drug for treating AIDS-related parasitic disease, and raised the price from USD 1,350 to USD 75,000 for a monthly course of treatment. Denouncements from individual doctors followed, then from associations, and finally the US congress and presidential candidates singled out Turing and Shkreli for critique.
High drug prices are not unusual in the US, however. The most recent top 10 list of drug prices I have seen posts two Hepatitis C treatments from Gilead Sciences at the very top, both above Daraprim. There are some differences that might explain this. First, Daraprim actually cost 13 times more in the US than in most other developed nations before the price increase, which increased the price by another 56 times. As a result, in the US patients pay 750 dollar for a pill that costs 1 dollar in Australia. Second, Daraprim is an old drug with no patent, so it could be made and sold generically – but Turing has distribution rights preventing that from happening. Gilead’s drugs are recently developed and under patent still.
But we should not make too complicated explanations of things that have simple reasons. In a recent paper in Administrative Science Quarterly, Sinziana Dorobantu, Witold Henisz, and Lite Nartey look at how society responds to controversies about firms, and they find very clear patterns. Moreover, they look at gold mining internationally, so their research is free of any specifics of the US and the pharmaceutical industry. Their explanation is simple and provocative: history and track record matter.
First, people form beliefs about firms based on how they have behaved in the past, and will lash out at firms with past misbehavior. Shkreli already had triggered a pricing controversy, so he definitely fit this pattern.  Second, the early statements from stakeholders set the tone for the rest, both because others often follow their assessment of an action as bad (or good), and because prominent stakeholders are imitated by others. Shkreli’s initial critics included Infectious Diseases Society of America, which is easily among the most prominent voices for this form of health care. Third, how people form beliefs about firms is largely a function of firm track records. Often a firm will have defenders who try to counter criticism, but friends are earned through actions. Martin Shkreli had no friends rising in defense, and no track record suggesting a reason for having any.
A true cynic might note that none of this matters, because getting 750 per pill is still a lot better than getting 13.50, or 1 dollar. But here the research by Dorobantu, Henisz, and Nartey shows that the cynic should be careful, because the stock market pays close attention to critics of the firm, and the firm loses value when it is under fire for misbehavior. In the case of Turing’s price increase, things went even further. Because drug pricing was getting legislative attention, the Nasdaq Biotech Index fell 4.4%, shrinking the value of an entire industry as a result of the actions of one company.

Why Aren’t Automakers Embracing Digital Business Models?

Below is my most recent post on Harvard Business Review:
BMW is one of the best car makers on the planet. It is also thinking seriously about what digital transformation means for the car business.
Its cars now have Connected Drive, a platform that allows drivers to purchase apps for traffic, messaging, and for starting the the engine from a distance. The new BMW is also packed with electronics that allow the user to experience different driving modes, from sporty to gas-saving, substantially changing the feeling of driving the car.
And yet BMW is still not making full use of digital business strategy – nor are any other car makers.
Consider: BMW charges €360 to unlock the ability to access the apps on the Connected Drive. Some apps (e.g. Remote Services) cost €80 and others (e.g. Real Time Traffic Information) can be rented for €45 over 6 months. If one spends a hefty amount of money on a new car, paying €80 or €45 for an app doesn’t seem too expensive, but needing to pay €360 to just activate the ability to download the apps seems totally wrong.
Contrast this with the approach taken by Apple. Making money on complementary products is one of the features of Apple’s business model. How does the model work? You sell the hardware and then you sell low priced apps (some of them are even free) to increase the value of the hardware. The apps represent a complement to a car and the Connected Drive is a store to sell complements, but why does the user need to pay to enter the store?
Imagine buying an iPad (especially in the early days of this product) and then having to pay €100 (or even €50) to access the App Store. This would have been a serious barrier. Following Apple’s logic would encourage BMW to make Connected Drive free, something that would make sense given the low marginal cost to BMW of doing so. The bigger lesson here is that you should always allow the customer a free entry into your digital store and then charge small amounts for the products sold there.
Here’s another way digital business principles might play out differently for BMW and other carmakers: renting engine capacity.
If you look under the hood of BMW’s Series-3 vehicles, for example, you can get horse power of either 110, 150, or 190, depending on whether you’ve purchased a 316, a 318, or a 320. However, you might be surprised to learn that BMW uses the same 4-cylinder engine in all 3 models, except the electronic components don’t allow the engine that is sold in the less expensive model to get to the higher levels of horse power.
Why couldn’t the company make a car that allows a driver to either upgrade or rent the engine power? Say you buy a 318 model with 150 horsepower for casual driving, but then you rent the 190 HP to go on a road trip? Alternatively, could you buy a car with 150 HP but after a 3-year period pay to unlock additional horsepower permanently? If the hardware is an issue, this unlocking could happen in a dealership.
We see these free-premium-rent models all the time in other digital businesses. When you download a fitness app, for example, you can try a free version first, and then can pay to unlock premium functionality later on. Or you can rent some functionality, such as a €9.99 a month subscription to an app that gives you a personalized training program.
When I talk to auto industry executives, the reason why they don’t want to systematically offer engine upgrades is that they want the customers to sell their old car and buy a new, more powerful car. Fair enough. But they may be missing out on both new customers and new revenue opportunities. Clearly, when commuting to work or driving in the French countryside, one doesn’t need 190 horsepower engine (not only because of the high fuel consumption, but because of the high probability of getting a speeding ticket). But on a vacation to Germany, where there are no speed limits on the autobahns, 190 horsepower could come in handy. As the car already has the different driving modes that are controlled electronically, it seems that the HP control is also possible.
There are also possibilities for automakers like BMW to combine user data, software upgrades, and digital business models to “nudge” customers to try new features they’ve not used before. Consider that your Connected Drive apps might know that you’re planning an upcoming trip to Cote d’Azur, where the speed limit is 130 kilometres per hour. The car itself could ask you if you’d like to implement a temporary, over-the-air engine upgrade.  Perhaps automakers could even offer a “vacation bundle” – additional traffic, weather, and events information along with an engine upgrade that lasts the length of your trip.
Tesla does now offer a self-proclaimed “ludicrous” mode upgrade to Model S that allows reducing the acceleration time of your car by 10%, and you don’t need to sell your old Tesla to get this upgrade. However, Tesla still asks you to buy the upgrade (about $10,000), not to rent it, although the rental of additional power should be at least technologically feasible.
Clearly, the makers of physical products (like cars or home appliances) understand that digital convergence is the next frontier. However, they often don’t look carefully or creatively at the business models this might inspire. The physical asset itself is just the beginning

Lean in or Lean out? Unfair Treatment Stops Women’s Careers in Many Ways

We have by now learnt a lot about how women’s careers are held back by unfair evaluation and promotion procedures, and it gets worse at higher levels in the organization. The glass ceiling exists at some point before the executive suite, unless we are talking about the more symbolic executive offices that are seen as good women placeholders. Women know this too. A centerpiece in the discussion about women’s careers is the book “Lean in: Women, Work, and the Will to Lead” by Cheryl Sandberg, COO of Facebook, which offers career advice for women to get ahead. Many women took the recipe-like advice as a way to behave more like men, in order to the get ahead the way men do. Others asked whether title “the Will to Lead” and its focus on women’s behaviors was a way of blaming the victims of a system set up to make them fail.
It is fair to say that the discussion of that book is a sideshow for most women with careers. They care about the hiring and promotion decisions that they are exposed to, and they doubt that these are fair. That makes sense: why should they be any different from the others?  Chances are that they have been hit by unfair promotion criteria at some point in their career.
Now research by Raina Brands and Isabel Fernandez-Mateo in Administrative Science Quarterly has revealed a cruel twist on this story. In turns out that people adapt their behaviors to the fairness of the system they are in. If they are treated fairly, they will reach for opportunities. If they are given signals that they belong in a group, they seek to join it. And once you think about those two mechanisms, it is obvious what happens to women seeking executive positions. They are not treated fairly and felt to belong, and the rejections from positions that they (often) should have gained discourages them from reaching for new opportunities. After all, who plays a losing hand? Naturally this accumulates over time, because more experience means more rejections, so exactly the women best placed to become executives are most likely to think they cannot reach that level.
This is not just a story about women. Unfair treatment can hold back a group in the short run. In the longer run it creates discouragement and resentment, and the members of the group starts holding themselves back. They are leaning out of the unfair

system, looking for better places to work. The labor market gets split as they avoid the career paths with unfair treatment, and organizations need to fill their positions from an increasingly narrow and less talented pool of applicants. The firm that shows through its hiring that it has a problem with female, black, Muslim, and Hispanic job applicants will learn the long-term consequences of narrow hiring.

Probing the Protests: Firms can learn to Avoid Activists

Often we see popular protests against firm initiatives. Recently the Standing Rock Sioux tribe and environmentalists organized protests against the planned Dakota Access Pipeline, which was scheduled to run through sacred grounds and across a river. The project has been suspended not because of protests, but because the Army Corps of Engineers blocked the measure needed for it to be legal. Could the construction backers have understood that the pipeline routing would lead to protests? In retrospect it seems obvious that an oil pipe through sacred land would be seen as a rough equivalent to an oil pipe through a church, and would lead to some anger. But the more general question is, can firms learn to avoid provoking activists?
It turns out there is research showing that such learning happens, at least for firms that are experienced with protests. An article in Administrative Science Quarterly by Lori Yue, Huggy Rao, and Paul Ingram studied the combination of two events: protests against Wal-Mart Inc. store development proposals and subsequent Target Corporation filing of store development proposals. This sounds complicated, but it is a really simple sequence. Walmart needs to file a proposal and have it approved in order to open a store (stores are big projects). After a proposal is filed, there can be protests against it (many dislike the idea of a nearby Walmart store), and Walmart can then decide whether to stop planning for the store. Walmart is known to file many proposals, and has a pattern of probing for places that are “protest-safe” by the seeing whether there is a protest or not.
But in our sequence, the next step is to see what Target does if there is a protest after Walmart’s filing. Here it gets interesting. For Target, it could be a simple rule to just avoid places with protests. In fact, they found that Target does avoid places with protests, but it was also learning in smarter ways. First, because Target knows that labor unions are uniformly unhappy with large low-price (and low-wage) department stores, it pays less attention to union-organized protests than to protests from other local groups. Second, it distinguished between protests that specifically paint Walmart as evil, versus those that are against any large store. Target is likely to enter when protests are specifically against Walmart, but to avoid places with protests against big stores. So, Target learns as much as possible from each protest.
And Target is even more sophisticated than what I just wrote. These learning patterns are what Target uses for locations that they are not familiar with, so they need to use protests to learn instead of relying on own local knowledge. If Target already has knowledge about a location, it ignores the protest and goes ahead with its own plans based on the commercial promise and its own assessment of risk.
Clearly, a corporation needs to be pretty unpopular (and to have unpopular peers) to become this good at learning from protests. And equally clearly, protests are not just temporary solutions, they are also signals that firms pay attention to and learn from. Protests have a deterrence effect, just as proposals have a probing goal.

Kickstarting the Disadvantaged: Activism in Venture Funding

Research and news tell the same story: there is discrimination both in employment and in business. Women are few and far between among executives and founders of technology firms, and claims of bias are often made, especially in Silicon Valley. On the financing side, venture capital firms appear to disadvantage women in executive roles. #AirbnbWhileBlack is hashtag collecting discussions of discrimination, and has led to Airbnb examining its processes for retaining hosts.
This looks like a problem for women seeking to start businesses, especially if those businesses are in industries with few women to begin with, like high technology. Even worse, the tendency to favor similar people to oneself – homophily – could make this even worse. Interestingly, a recent article in Administrative Science Quarterly by Jason Greenberg and Ethan Mollick has found a counter effect. The idea is that if a minority thinks that it is discriminated against, it will be especially supportive of its own members. It will not only favor its own, as all groups do, but it will do so in an activist way. If this happens, being recognized as a disadvantaged minority – like women in technology – will lead to better treatment, at least from members of the same minority.
Does it happen? It is not clear whether this is always true, but one good place to look is in crowdfunding, where ventures and their founders are presented to a “crowd” of any interested funder, and they in turn decide what ventures to back. And indeed, women Greenberg and Mollick found that women targeted women’s ventures for funding, and did so

especially for industries were women are known to be scarce. So, women especially supported other women not fashion or publishing, where women are frequent business founders, but technology, where they are scarce.

This is clearly not a reason to think that discrimination will balance out. Crowdfunding is the form of funding where this type of activist support is most effective, but most venture funding is not down through crowds – and we already know that venture capital firms, for example, have mostly male executives. Also, activist funding does not have large effects when there are few women funders to begin with. So, we can conclude that this provides some relief, but it is a less fair solution than simply evaluating ventures on their merit.

Getting the Orange out of My Head: How Respect can set Inmates Free

Organizations can have very different work environments, including differences in the respect they give to employees. Organizational cultures differ, and managers differ, in whether they see employees and valuable and how well they acknowledge this. Many organizations think that it makes a difference – notice how I used the word “employee” just now, but actually words like “colleague” and “team member” are frequent in actual work. Does this matter?
For an example of how much this can matter – in a very special context – a recent paper in Administrative Science Quarterly by Kristie Rogers, Kevin Corley, and Blake Ashforth looked at an organization that operates professional call centers as part-time work for selected inmates in prisons. Every day inmates go to work in their orange jump suits (yes, just like in the TV series). Every day they go back to the prison wing after working. But this work is not like the demeaning chain gangs that we see in some old movies; the organization (Televerde) values its inmate workers and gives them both encouragement and respect.
So what happens? The respect they get from their Televerde managers, and from customers, changes lives. They get a specialized respect based on the value of the work they do, and their performance, and this gives excitement and self-respect. They get general respect from being seen as real people with lives and accomplishments, not inmates with orange jumpsuits and numbers, and this gives ideas of a changed and improved life. Together, these two kinds of respect, and especially the general one, puts the inmate-workers on a path toward removing themselves from their identities as current and future inmates, and attaching themselves to a new identity as a professional doing legal and respected work out of prison.
It happens impressively fast. These changes were easy to see over a period of less than a year (for most it was much faster), even though the Televerde workers were still in the prison wing, with their old friends and controlling prison wardens, every day after work. As part of the identity journey, they needed to transition from their old thinking habits – the orange in their heads – to a new way of seeing themselves as part of a regular civil society that they could not yet reach because it was outside the prison walls. Remarkably, they were able to not just see their inmate identity and their worker identity as separate beings coexisting in their minds, they also could shift to a new and holistic identity that would guide their lives after they were released from prison.
Giving workers respect is seen as important also in regular organizations, with no inmate workers, but there is a certain degree of cynicism about its effect, and there are also managers who don’t think it matters. After seeing how transformative it can be under these conditions, when it is done honestly, maybe it is time to reconsider.