This is the first of a two part series on non-hierarchy and consensus decision making. In this first post I’ll be examining the differences between hierarchies and non-hierarchies and discussing some of their pros and cons.
In my next post I’ll delve in depth into how non-hierarchies work and some of the observed pitfalls I’ve found.
I have been thinking a lot of late about non-hierarchy. Since April I have been involved in Extinction Rebellion; the decentralized non-hierarchical environmental direct-action group founded in the UK last year. As a result I have been re-assessing my experience of non-hierarchy in Exchange Dublin and considering what lessons I took from that experience.
Exchange Dublin was a non-commercial all-ages collective Arts space in Templebar. It operated between 2009 and 2014. It was a non-hierarchy. Every week there was a collective meeting where anyone who wandered in off the street could have a say in how the space ran.
Exchange Dublin was a small 5 room complex. There was a kitchen / reception, a gallery space, an event space, a bathroom and a small office / store room. The group that ran the space effectively decided what should happen in those rooms every week.
All of the decisions in Exchange were made by consensus. There were no employees, only volunteers, and the crew of people involved changed over time. Exchange Dublin operated using what I will call the ‘strong-block’ model of consensus decision making; although it was rarely used, anyone could veto any proposal.
I was active in the space for around 3 of it’s 5 year existence and perhaps more than any other group Exchange Dublin changed my life. When I walked into Exchange I had never encountered a formal non-hierarchy before, but over those 3 years of involvement and facilitation I learned an enormous respect for the systems and dynamics that allowed it to operate.
Over the past decade non-hierarchy and consensus decision making have begun to spread. Much of this is thanks to the role they played in the Occupy and Indignados movements and their role in spreading them to a wider audience. Today both are utilised by a great many newer and less formal organisations and coalitions.
I am not someone who focuses on a single big idea. I’m more of Isaiah Berlin’s fox than his hedgehog, but the closest thing have to any over-riding belief is my belief in non-hierarchy. Non-hierarchy is how I want to run my business, how I believe groups can best utilize their members and how they can best serve the common interest.
This is not to say that non-hierarchy is easy, or easier, than hierarchy; In fact I would argue that it is harder. It is also not to say that Exchange Dublin didn’t have problems, it did; many of them and some quite serious. However with the caveats I discuss below and with the recommendations I lay out in my next post I believe the investment in non-hierarchy leads to better outcomes.
Before I begin comparing and contrasting hierarchy and non-hierarchy it is worth defining both terms as I understand them.
A hierarchy (from the Greek hierarkhia, “rule of a high priest”) is an arrangement of items in which the items are represented as being “above”, “below”, or “at the same level as” one another.
A hierarchy is any group of people where one portion of the group has power over, or more power than, another portion of the group.
Most of the organisations you will have encountered are probably hierarchies. In modern societies hierarchy is the default mode of organisation. Concretely, any organisation with a manager or ‘leader’ is a hierarchy.
Hierarchies, particularly for larger organisations, tend to have multiple levels of hierarchy. They tend to centralize decision making power, and decisions tend to flow down from the top to the bottom.
Non-hierarchy: lacking hierarchical structure; egalitarian.
Non-hierarchy is a system diametrically opposed to hierarchy, and indeed my definition of non-hierarchy is simple:
A non-hierarchy is any group where all members of that group have equal decision making power.
In non-hierarchies there is no high priest; no-one is above anyone else. In the most successful non-hierarchies individuals have almost complete autonomy to act within the structure of the organisation. This autonomy is often expressed with a principle of ‘forgiveness rather than permission’ although more formalized non-hierarchies tend to limit this, sometimes with boundaries around areas where permission is required.
Non-hierarchies tend to be run by consensus; in fact to remain non-hierarchies it is almost essential that they are run by consensus. Consensus decision making is where all members of the group agree to, or at least do not disagree with, a decision. I will cover consensus decision making in some detail in the next blog post.
As everyone reading this will have encountered a hierarchy, and because a lot has been written about how hierarchies operate, I will not to delve too deeply into how hierarchies work. Before I discuss the strengths and weaknesses of hierarchy, I would like to introduce a small tangent on different forms hierarchy can take.
Hierarchies can be explicit or implicit, formal or informal.
In an explicit hierarchy the power relations between different members of the group are obvious to all members of the group. Everyone can determine where they and others sit within the hierarchy. The means of determining the hierarchy is usually clear; everyone understands the logic of why someone is above or below someone else, even when that logic is arbitrary (more on this later). A teacher / student relationship is a good example of an explicit hierarchy.
In implicit hierarchies power relationships are determined by criteria that are not clear to all members. The hierarchy may change depending on various social or political factors and even those factors may change. In implicit hierarchies it is sometimes only possible to determine where you sit in the hierarchy in relation to another member of the group. It may not be possible to see where another member sits, either absolutely or sometimes relatively. In rare circumstances it may not even be possible to determine where you are placed in an implicit hierarchy relative to another member of the hierarchy.
Many social clubs and organising committees are implicit hierarchies. I would argue that any group where politicking takes place is an implicit hierarchy.
Hierarchies can be either formal or informal. In formal hierarchies the rules determining the hierarchy are codified, in informal hierarchies the rules are not codified. Most formal hierarchies are explicit and most implicit hierarchies are informal; but there are exceptions.
There are also hybrid hierarchies. Hybrid hierarchies have mixed explicit / implicit and formal / informal components. Many, if not most, hierarchies are mixed.
Hierarchies have some advantages. Notably, they are easy to set up and run. You establish the organisation, install yourself at it’s head and tell anyone else you can convince to join what to do. There are plenty of examples of hierarchies all around you to copy and learn from and people who join the group will know the role expected of them.
The tendency for those who establish the hierarchy to place themselves at the top of it is widespread. This appeals to the founder’s ego and generally results in the largest share of the organisation’s surplus (be it financial or social) flowing to the creators of the hierarchy. This clearly is the key benefit to those creating the org; they will receive the largest share of the benefit from it if it is a hierarchy. This is also the reason why hierarchy is so prevalent, it is perceived as being in the self-interest of the founder to create a hierarchy with themselves at the top.
Hierarchies tend to be stable in the sense that organisations with hierarchies can last for centuries, well beyond the involvement or lifespan of it’s founders.
In the specific case of decision making, especially in strategic decision making, hierarchies are faster than non-hierarchies. This is because a very small group of people, sometimes one person, can make a decision. However as I discuss below, making the decision and implementing a decision are different faculties.
Hierarchies also remove the expectation of thought from most people in an organisation. While this has a strong negative effect, it also has a positive effect; it can reduce the cognitive burden of being involved in the organisation.
Many of the behaviours I discuss below are emergent properties of hierarchies, but that is not to say that all hierarchies will exhibit all of these characteristics. Many hierarchies actively work to reduce some of these issues. I would argue however that if they wish to avoid them all hierarchies must implement strategies to prevent them emerging, and I do not believe any hierarchy can avoid all of these problems and remain a hierarchy.
Most hierarchies disempower most people in an organisation. Hierarchies are generally pyramidical, with a smaller number of people with more power at each level as you work your way to the top. This means most people aren’t, or don’t feel like they are, in a position to shape the organisation. Consequently there is a tendency for people to be unhappy or at least ambivalent towards the organisations’s strategy.
Hierarchies tend to give more social status, power and income to people depending on where they are in the hierarchy. As these are motivations for people there is also a tendency for members to focus on climbing the ladder and playing politics rather than the success of the organisation. Ultimately this is how hierarchies self-perpetuate; it is always in the interests of the people moving up the hierarchy to take the place of those above them.
The imbalance of power tends to lead to less than optimally efficient and happy groups. Only a very select number of people at the top of the hierarchy have attained all they can, everyone else has less. Even when neither power nor wealth is a motivating factor, many people will feel less than fulfilled in roles where they cannot maximally shape an organisation.
Frequently the desire to move up the hierarchy, for self actualisation or extrinsic factors, will lead people to leave the organisation in search of somewhere where they feel they can have a bigger impact or gain greater rewards. This leads to organisational turn-over; which is a problem in many hierarchies.
Another disadvantage of hierarchy is that people feel threatened by those lower down on the ladder, who they perceive are trying to climb up. This can manifest in several ways; it can lead to active sabotage by those above of those below, it can lead to an environment with reduced trust and it can result in responsibility and power being concentrated into key people in a hierarchy, to ensure that they are the ones who benefit from success.
As a consequence many individuals in hierarchies adopt a strategy of favour seeking from those above them; both to maintain their position but also as a protection against that jealousy. This boot-licking behaviour is often seen as unseemly by everyone involved, but is nevertheless also quite widespread.
Formal hierarchies often have formal approvals processes for new ideas or decisions. This generally involves sign off, or agreement from people higher up a hierarchy for ideas coming from lower down. This can result in a painful slowdown. People wait for the manager to make a call on something, or delay action until they have cleared it with someone above them in the hierarchy. Often politicking, or even just the changes made to an idea passed through several hands, waters down it’s effectiveness.
Even when such formal structures for approval don’t exist, the powerful social effect of hierarchy often does the same job; indeed the reluctance to “step on someone’s toes” is most pronounced in informal hierarchies. If it can’t be determined who exactly is responsible for what then people tend towards inaction.
Humans mostly avoid conflict. We tend to be risk adverse and we have a strong status quo bias. This means that many of us are unwilling to challenge authority and even when we are we often seek approval for our actions. The next person up the ladder is as likely to feel the need for approval as the first; so decisions tend to be passed up the chain, which slows the process down.
I’ll refer to this process as bottle-necking, and it is most pronounced when one person has oversight over too many decisions, or when too many people are involved in a tough decision.
To some extent the problems of approval seeking and bottle-necking also effect non-hierarchies but the effect is more pronounced in hierarchies; because of the shrinking number of people capable of making a decision on each level.
Silo-ing is another consequence of the incentives model in hierarchies. Silo-ing is where one group claims ownership of a particular scope of activity, and then makes decisions based only on that scope, without reference or consideration for the rest of the organisation. It usually promotes the value of that scope above other parts of the org, thus creating an informal hierarchy, often within a formal hierarchy. Silo-ing tends to strengthen loyalty within the subgroup at the expense of trust in the organisation.
A common complaint of hierarchy is that at each level people are further removed from the ‘coalface’. They do not know the details of the problems they are making decisions over. That this leads to poorer decisions goes without saying, but often it also leads to frustration amongst those at the coal face, who can see the problems and often the solutions, but cannot change the organisation.
Oddly the issue of experience and information cuts both ways in hierarchy as there is also a problem with information being passed down from the top. This can be seen very often for decisions and outcomes, but perhaps more importantly frequently happens with the rationale for decisions.
Organisations thrive on ‘buy-in’ (sometimes referred to as alignment) or the belief that all the members of an organisation have in it’s goals and strategy. A lack of buy-in usually means unproductive members and half-hearted attempts to fulfill strategy. People tend to need intrinsic motivation to do their best work, and buy-in is essential to that. The most common reason a project fails is because people did not believe it had to succeed.
By shrinking the number of people involved in a decision, and not sharing the rationale for that decision hierarchies make it very difficult to achieve buy-in. With reduced context people who have to implement the strategies passed down to them will doubt their value; people tend to dispute a decision they cannot see the logic of.
Concentrating decisions into a smaller and smaller number of people may also tend to lead to worse decisions. Hierarchies miss the opportunity to use the wisdom of crowds to develop the best solution to a problem.
The most common place you will encounter non-hierarchies are in informal social groups but an increasing number of companies and organisations have adopted non-hierarchy or flat practices.
In many groups of friends, no-one in the group is more important than anyone else, everyone has the same decision making power and everyone in the group is valued equally.
The growing movement for non-hierarchy in business is often referred to as ‘flat organising’; as the companies who employ it don’t follow the pyramid structure hierarchical organisations use.
As with hierarchies there are formal and informal non-hierarchies. Formal hierarchies have rules about participation, decision making and, perhaps especially, conflict resolution. The rules tend to maintain the egalitarian nature of the group and they are designed to limit the power of individuals within it.
Informal non-hierarchies tend to be small groups of friends, very small companies and business partnerships. There are no rules about group dynamics and social pressure is the only force that maintains the non-hierarchy.
Often the rules formal non-hierarchies operate under are emergent from an informal non-hierarchy. As non-hierarchies grow they tend to formalize.
All formal non-hierarchies are explicit; it is clear to everyone that they are organised non-hierarchically. Informal non-hierarchies are often implicit; it is rarely acknowledged that there is no hierarchy, often because it doesn’t occur to anyone that there shoud be a hierarchy.
I am far more interested in how non-hierarchies work than hierarchies, and as they are less well understood I will try to give a brief outline of how they operate.
This is a brief overview, the entirety of my next blog post will be about organising and operate non-hierarchies and what pit-falls they face. All of these musings are heavily skewed by my personal experiences with non-hierarchy; these are definitely not the only way non-hierarchies can be organised, and they may not be the best ways.
In non-hierarchies there are no leaders and there is little to no people-management. The group decide amongst themselves what needs to be done and who will do it. Often decisions are made by consensus, occasionally by voting. I will argue that for non-hierarchies to remain so, they need to use consensus to make decisions.
Consensus decision making essentially means that everyone involved in making a decision agrees with, or at least does not object to, the decision.
As everyone in the group has the same power, anyone can propose an idea, whether the group follows through and takes up that idea is down to how good the idea is and, often, how well argued it is by it’s proponent.
In a typical consensus decision making system a decision is made after some debate on a proposal. There are 3 ways a proposal can go:
- The proposal is accepted as is.
- A modification or counter proposal is made.
- Someone blocks the proposal.
If the proposal is accepted it means that at the end of the debate no-one is willing to block it. A block is a veto that every member of the group has. Generally blocks are few in number and counter proposals are much more usual.
A proposal may be accepted without the ringing endorsement of a large portion of the group. People may not feel strongly enough about a proposal either to offer to help implement it, or to block other people doing it. This generally means that a small sub-group will attempt to carry out the proposal; often this group are the proposers.
There are a few different ways that groups decide how to distribute tasks, and again I’ll go into this in some detail in my next post, but often the group predetermines roles that people take on ahead of work and then do the work required by that role, occasionally work is volunteered for on a case by case basis, although in some circumstances it can be assigned by the group. I have also heard of groups trying more experimental methods for assigning work; like lotteries.
The most significant advantage to non-hierarchy is that anyone in the group can shape the group to the same extent as anyone else in the group. This tends to maximize each member’s utility. No-one feels that there are arbitrary limits on how much they can achieve and this directly leads to happier and more productive organisations.
Knowing that they can be involved in any decisions they feel strongly about members have much stronger buy-in for group decisions. When someone is involved in the decision making process they get to see the rationale for the final choice. Thus, even when that choice wasn’t the option they initially were in favour of, they tend to respect and implement that decision.
To some extent buy-in is improved even when a person is not directly involved in making a decision; over time people trust the process and understand that their peers tend to make the right decisions. They also know that they had the opportunity to be involved in any decision that was made.
As there is no extra power to be gained through manoeuvring there is almost no focus on politicking or self-advancement in non-hierarchies. This leads to more of an individual’s energy being spent on the group’s aims and the maintenance of the group.
People who would in a hierarchy be concerned with climbing the ladder or impressing an individual above them, instead focus on the activities that they believe will strengthen the org.
While, as I discuss below, strategic decisions are slower in consensus groups implementation can be much faster. This is because the people closest to the coal-face understand what has been decided and as a result can immediately start implementing a decision. There is less room for misinterpretation than there would be as information is passed down a pyramid, and the individual autonomy members of the group have allows them to focus on priorities they identify to drive a decision.
This reduced distance to action, where people can get as involved as they want as quickly as they want, is one of the real strengths of non-hierarchy. It allows super-enthusiastic newcomers to get involved straight away and continue to drive the organisation forward with an energy that hierarchies struggle to compete with.
The autonomy to act also reduces bottle-necking. Once a decision is made individuals or sub-groups have a mandate to act, and while operating in that decision space shouldn’t feel they need more permission. As a result less time is wasted waiting for other people to decide something.
This advantage can be over-stressed however. Non-hierarchy does not remove bottle-necking; usually in non-hierarchies there are still individuals with specialist information and even autonomous actors will need to wait for some input from them.
As everyone is an equal member of the group the desire to place a sub-group’s aims over the main group’s aims are reduced. This is especially true if each sub-group has open membership.
In turn this can lead to reduced resource competition between functions of the organisation. Different sections are more awake to the idea that what they are working on may not be the area the most resources should be expended on.
Another sizable advantage of non-hierarchies is that decisions tend to be made by larger groups with more input from people with wider lived experiences. This allows non-hierarchies to use the “wisdom of crowds” to make better decisions. The wisdom of crowds is a widely observed phenomenon where the average decision of a group is better than the decision of any individual within the group.
However this only tends to apply in groups where there is some disagreement over decisions or direction. If non-hierarchies fall into a pattern of close agreement between members it can end up with “groupthink”; where much worse decisions are made.
A function of the extra time non-hierarchies spend on deliberative decision making is also that those decisions may be better than decisions taken in a hurry by an individual or small group in a hierarchy. Swift decisions are not always the best decisions.
Non-hierarchies rely on the autonomy of the group members. For newer members who have not experienced non-hierarchies before, this can initially be a drawback. However once group participants get past the point where they start putting themselves forward there is a significant organisational advantage; far less group effort is spent on people management.
This means that a non-hierarchy of the same size as a hierarchy is more productive; there are more people doing work on the coalface, and fewer or none, spending their time telling people what to do.
Additionally, experienced or exceptional individuals can continue to spend time on tasks they are particularly good at, or enjoy, because there is never a point where they are “promoted” away from that task.
Equally importantly, when someone gets fed up with the role they are currently doing there is generally no barrier to changing to another role within the organisation. This can reduce org turnover as people move on to other things within the group, rather than an entirely new organisation.
Despite their many advantages there are several disadvantages to non-hierarchies.
It is more or less impossible to have a functional non-hierarchy with members who have competing or conflicting aims. As mentioned above and explored in detail in my next post; non-hierarchies need to use consensus decision making to remain non-hierarchies.
Consensus decision making in it’s purest form requires agreement amongst all participants in a group. At the very least it means that no-one in the group is strongly against an action or activity. In a group where everyone is strongly aligned on aims and agrees with the rules, disagreement tends to be about strategy and tactics, and most of those disputes, while sometimes difficult to resolve, are resolvable.
However, groups with competing interests can never reach an agreement on a zero-sum issue; such as resource division or an issue of censure. As a result a decision that illicits a conflict will hamstring the organisation.
If this happens enough times the group will factionalize and a hierarchy will emerge that can result in other decisions being blocked; even when those decisions are not zero-sum.
An oft-cited criticism of consensus is the UN problem. In the UN the security council operates on the basis of consensus decision making. However it’s effectiveness is often hamstrung, especially on tougher decisions, by competing interests.
In general having strong aims and rules reduces this friction; particularly if those aims and rules shape who is allowed into the organisation.
While implementation time on non-hierarchies can be swift, decision making can be slow. This tends to be particularly true of strategic decisions. A deliberative process requires deliberation and almost all of the advantages non-hierarchies have come from widening the decision making pool.
As a result the key trade off with non-hierarchies is the time required to make decisions. This can sometimes lead to groups feeling like they spend all their time discussing without acting.
There are many ways to reduce this issue, and I will explore some in detail, but this problem should neither be under-estimated nor over-stressed.
In groups where a number of people are contributing only a small amount of time, if a significant portion of that time is spent on deliberation then the group will get nothing done. Conversely the extra time spent on decision making is rarely regretted in groups where individuals are contributing much more of their time to the organisation.
As mentioned, non-hierarchies tend to find it harder to make strategic decisions. This is largely because those decisions have the ability to effect everyone in an organisation. It’s often crucial that everyone understand the strategic rational and be a part of the decision.
This however requires everyone in the org to be notified in advance of a key decision and be provided the opportunity to contribute to it. Obviously this slows down the process even further. Not only will non-hierarchies spend longer deliberating a key decision they also need to postpone that deliberation.
Depending on how many people are involved in the organisation, this can result into a slowdown on certain decisions of days, weeks or conceivably months.
Another consequence of consensus decision making, particularly the strong-block model, is that the status quo is heavily favoured. As only new proposals can be blocked, any block results in the status quo being maintained.
While in my experience blocks are rare; when they occur the org always continues as it previously had. This can be disastrous in certain circumstances; such as a tough decision with no ideal solution.
Non-hierarchies do not remain non-hierarchies without the people who make up the organisation spending effort to maintain them. Non-hierarchies require people to relinquish power, and to ensure that no group gains power at the expense of another.
The power that needs to be relinquished is always the power you yourself possess as a member of the group; this means that everyone needs to make an effort to keep a hierarchy from emerging. This takes significant buy-in. Over the years I’ve come to the conclusion that a non-hierarchy can only be maintained if the people involved in it believe in non-hierarchy.
As a result non-hierarchies are vulnerable to bad actors and as result need to protect themselves from them from the get go.
Flat groups need to constantly check that a hierarchy has not emerged, and make sure through constant effort and self appraisal that they are being honest about the ability of anyone in the group to get involved in something or take individual action. Deciding to remove a formal hierarchy does not make an organisation a non-hierarchy; only if each member of the organisation can have have the same impact as anyone else has non-hierarchy been achieved.
The added effort from everyone involved to maintain a non-hierarchy can in some ways offset the productivity advantage non-hierarchies have from having fewer people managers.
Non-hierarchies have a further problem with evaluation of actions by individuals. In groups with certain motivations people will tend to take on only certain kinds of work. This can result in most people doing actions they see as particularly valuable, or at it’s worst, most enjoyable.
As individuals have increased autonomy in non-hierarchies they have greater scope for pursuing goals they value higher; however this means there is a greater requirement for individuals to accurately assess what needs to be done.
This was a problem we saw to some extent in Exchange where there were jobs that needed doing but that few people volunteered for. While we were able to overcome this on a case by case basis with a rota once we had identified the gaps; the larger problem is actually about identifying the gaps in the first place.
In a system where all members have equal power, accountability and dispute resolution can be pain points. These can both be seen as zero-sum and consensus decision making struggles to deal with these effectively.
On accountability, verifying that someone has done work they undertook can feel like management or lack of trust. When an issue _is_ discovered censuring someone can be much harder to do than in a hierarchy.
Dispute resolution can be even worse as pitting individuals against one another can lead to factionalizing, and it is very hard for a group to make a dispassionate decision.
Both of these issues we encountered in Exchange; and we had no systems in place to deal with them. My learning from that experience is that some system needs to be in place right from the start, and anyone involved in the group needs to accept the system for resolution.
As touched on previously, the best way for non-hierarchies to resolve conflict is to formalize with strong rules. However, this only works if the rules are agreed in advance; generally amongst a small group of founders who have no conflicts at the time they are agreed.
This places an onus on founders of non-hierarchies to get rules for dispute resolution right from the off. While it is possible to add rules later, it will be very hard to add rules around an issue that has already been disputed. This is because in many circumstances turning a decision from a conflict into a rule can alienate people involved in that conflict.
One of the largest disadvantages of non-hierarchies is a heavy requirement for self-motivated people who then must be prevented from taking over. The reliance on self-motivation is particularly pronounced in the initial stages of a non-hierarchy when people are reluctant to ask each other to do things. The only way things get done is for people to take them upon themselves.
This can be mitigated to a certain degree, but for a non-hierarchy to remain a non-hierarchy those people who throw themselves into things need to be discouraged, either by themselves or others, from taking on too much work. On the flip-side people who generally need to be told to do things need to be encouraged to take on work themselves.
As a concrete rule, any system without leaders requires everyone to show leadership.
Lastly in groups using consensus there is also a bias in favour of people who are good at arguing in groups and are not intimidated by speaking in public. While there are many ways to overcome this it is still a problem that needs to be overcome for a hierarchy of out-going or self-confident people not to emerge.
I want to focus in my next post entirely on the operations of non-hierarchies, but there is a last issue I would like to touch on in this post; namely the moral argument against hierarchy.
All hierarchy is arbitrary. If we accept the proposition that all people are born free and equal, then nothing makes any individual inherently more deserving than another of higher status.
Thus a hierarchy based on birth or birthright is an arbitrary determinant of status. Just as the son should not be punished for the sins of the father neither should the daughter be rewarded for the deeds of the mother. We enter this universe a random chance at the end of an unimaginable sequence of random chances - and we are no more worthy than any of the other possibilities that could have taken our place.
Likewise, age and seniority are similar accidents of causality. When you were born was a random thing. The fact that you were born before or after your sister was not pre-determined. Extending the thought process back through the generations, your race and perhaps even your species are due to the random fluctuations of chance.
Your education, and the skills you acquire throughout your life are also down to chance. Where you were born, in what era, to what parents with what resources. All the random chances that led your teachers arriving at your school, and all of the random events that needed to happen to impart the information you’ve acquired to you.
There is a pernicious myth about hierarchy; that those with the best skills will rise to the top. The idea is that hierarchies promote the brightest stars to the top of the chain, but it is a myth. The only skill that promotion to the top of a hierarchy will strongly correlate with is a skill at navigating hierarchies.
Those individuals who are best at the social and political interactions necessary to climb the ladder will reach the top; but it is unlikely that those are the skills most valuable to the organisation as a whole.