Beyond price: Trust me, I’m commercial

BEYOND PRICE: TRUST ME, I'M COMMERCIAL

Professor Michael Mainelli

Good evening Ladies and Gentlemen. Tonight we do have a lecture on 'Beyond Price' but first, trust me, we need a commercial. I'm pleased to advertise three forthcoming events. First, there is a symposium related to tonight's lecture, 'What Makes A Good Auditor?' here at Barnard's Inn Hall next week on the afternoon of 12 March 2007. Second, the London Accord annual conference at the Guildhall is on the afternoon of 24 March 2007. This year's London Accord theme is 'Investment Opportunities For A Low Carbon World'. Third, my next Commerce lecture will be 'What Makes A Successful Global Financial Centre?' at Allen & Overy's offices in the Docklands at 12:30 on Thursday, 2 April 2009. Admission to all three events is free, but reservations are required, available via the Gresham College website.

An aside to Securities and Investment Institute, Association of Chartered Certified Accountants and other Continuing Professional Development attendees, please be sure to see Geoff or Dawn at the end of the lecture to record your CPD points or obtain a Certificate of Attendance from Gresham College.

Well, as we say in Commerce - 'To Business'.

Trust Is Not A Good

[SLIDE: BOND, COMMERCIAL BOND]

The theme of tonight's talk is the role of trust in commerce. Trust: it takes years to build it, and a few seconds to destroy it. Examples abound of trust destroyed: Union Carbide and the Bhopal disaster, Exxon and the Valdez oil spill, Perrier and water contamination, Royal Bank of Scotland and the finances of the UK. Further, lack of trust has been the core of many project disasters, for example the 1992 London Ambulance Service Computer Aided Despatch project recovery in which I was involved where management, staff and suppliers lacked trust. More positively, the London Stock Exchange is still proud of receiving its own Coat of Arms in 1923, with the motto 'Dictum Meum Pactum' (My Word is My Bond). There are heaps of tongue-in-cheek Trust Me T-shirts - Trust Me: I'm A Doctor, Trust Me: I'm A Scientist, Trust Me: I'm An Economist, Trust Me: I'm A Lawyer, Trust Me: I'm A Politician, or Trust Me: I'm A Thief, and these days Trust Me: I'm A Banker.

Trust is reliance on other people. We distinguish trust from cooperation. Cooperation is when people agree to work together. Without cooperation society could not function at all. Without cooperation everything is a zero-sum game; no one benefits from helping others. All markets are based on confidence. Yet investors must take risks, new products and services must be explored, one needs to trade with strangers. When people rely on others they take a risk.

A cursory review of the etymology and definitions of trust throws up many many contexts. The Old English 'tr'owe' and Old Norse 'troust' are both about faithfulness. Trust as a noun is defined as 'a relationship of reliance', but also a dependence on some future contingency, credit (buying something on trust), holding property on behalf of another, as well as a charge or duty, for example a child committed to your trust. Further, as a noun it often means a legal or business structure for third-party beneficiaries, but equally a cartel to reduce competition, e.g. anti-trust laws. As a verb, trust means relying on someone or something, but also to entrust into care, believe someone or extend credit. And then there are the adjectival forms of trust.

Trust is more subtle than cooperation. Trust is when we leverage on a history of relationships to extend credit and benefit-of-the-doubt to someone. In risky environments trust enables cooperation and permits voluntary participation in mutually beneficial transactions which are otherwise costly to enforce or cannot be enforced by third parties. [Braynov and Sandholm, 1999] By taking a risk on trust, we increase the amount of cooperation throughout society while simultaneously reducing the costs, unless we are wronged. Trust is not a simple concept, nor is it necessarily an unmitigated good. Strangely, tonight might best start with a joke that exploits some of the complexities of trust:

[SLIDE: MOTHERLY TRUST]

Once upon a time there was a son with a very strict, but wise, mother who believed in the sanctity of marriage. The son invited his mother to dinner with his flatmate, the stunning Rosalinda. Mother noticed how beautiful Rosalinda was, the sly glances between her son and Rosalinda, and the adjacent sleeping quarters, so her suspicions were aroused. The son noticed his mother's unease and reassured her, 'Mom, really, Rosalinda and I are just flatmates'.

A few nights later, Rosalinda mentioned to the son that her beautiful family heirloom, a silver candlestick, was missing - rather pointedly noting that the last time she'd seen it, his mother had been to dinner. So the son wrote a letter to his mother: 'Dear Mom, it's not that I don't trust you and I'm not saying you 'did' take a silver candlestick, and I'm not saying you 'did not' take a silver candlestick, but one can't help but note that a silver candlestick has been missing from our flat ever since you were here for dinner.'

By return post the son received this letter from his mother: 'Dear Son, it's not that I don't trust you and I'm not saying that you 'do' sleep with Rosalinda, and I'm not saying that you 'do not' sleep with Rosalinda, but one can't help but note that if she was sleeping in her own bed, she ought to have found the candlestick by now!'

Our first observation must be that trust is about much more than just money; it's about human relationships, obligations, experiences, and about anticipating what other people will do.

[SLIDE: AGENCY THEORY]

With a concept so all-encompassing, it might help to examine trust diagrammatically. Trust is about reliance on others. We've explored commercial reliance in a previous lecture covering Principal-Agent Theory, 'What I Like About This Country Is That It Has A Nice Level Of Corruption' (4 October 2007). I'm flashing the slide before you for reference.

[SLIDE: INFORMATION THEORY]

Trust is also about relationships, communication and miscommunication, which we explored in a lecture on Information Theory, 'It's A Mad, Bad, Wonderful World: A Celebration Of Commercial Diversity' (17 November 2008). Again, I'm just flashing this diagram before you to refresh your memory.

Trust In Theory And Diagrams

[SLIDE: TRUST THEORY]

Our core diagram has two people. One is the principal; the other is the agent. The principal is the one who will depend upon the agent to do something. This dependence or reliance can either be involuntary - the principal has no choice - or voluntary - the principal places trust in, or obligates, the agent to perform something. Of course the situation is richer. The agent is often a principal too, while the principal is often partially an agent. If you find this confusing, just think of picking up a newspaper before buying it from a street vendor. You, buying the newspaper, are the principal. You obligate the vendor to sell you a readable, complete newspaper. However, the vendor is, for the split second you hold the paper before paying, also a principal. He or she has extended trust that you will pay. If you pay a lot of money in advance for something, say an airline flight, it's fairly clear you are the mostly the principal and the airline is mostly the agent. If you are in a messy contract to build an innovative sports centre with all sorts of adjusted stage payments, then your role as agent may be confused by clients' dreams, architects' plans and many other factors. [1] Reliance establishes the relationship of trust.

However, evaluation of this reliance will be based on two human factors. [2] First, the principal has an expectation or anticipation of the performance. Likewise, agents believe they understand the obligation and the anticipation of performance when they make decisions about how well they intend to perform. Second [3] the experience of the performance compared with the anticipation forms the evaluation. You don't need some deep mathematical understanding of Information Theory to realise that this trust system has great scope for miscommunication and mistrust. There are information cycles from the principal to the agent, and from the agent to the principal, with roles for senders, messengers and receivers. There is a feed-back loop of experience and a feed-forward loop of anticipation. Each loop readjusts expectation. Trust is a human system with feed-through loops, that is people's perceptions change anticipations. Thus, a system of trust exhibits bubbles and fat-tails of anticipation which in turn drive evaluations and future decisions, as well as fads and fashions.

If you find this too abstract, ponder briefly brands and trust. You see that other people have great trust in something, say a Harley-Davidson motorbike or Caterpillar construction machinery. Suddenly they're buying unrelated apparel or footwear based on the reputation established over repeated feed-through loops. Is that not a bubble underway? Further, there are informational cascades and herd behaviours, as well as situations where people adjust their anticipation post hoc. If everyone else liked something, perhaps I did too but failed to notice my delight at the time. A bit like Hans Christian Andersen's story 'The Emperor's New Clothes'.

[4] There are trusted third parties (TTP) such as trade financiers who guarantee payment upon successful shipment. Trusted third parties take many forms, bankers, insurers, guarantors. [5] There are also various forms of security, ranging from putting up money in advance through escrow accounts of various forms, where money or deeds or software are held in trust. Mankind has tried many mechanisms to establish trust. Finally, it is worth noting the difference between a 'one-off' transaction and a continuing relationship. Jenny Rayner defines reputation as 'A collection of perceptions and opinions, past and present, about an organisation which resides in the consciousness of its stakeholders'. [6] Reputation is formed during multiple rounds of interactions.

[SLIDE: NOBODY LOVES ME LIKE I LOVE ME]

Self-interest is a common feature in establishing trust. Tonight I promised that we would move at some point, 'beyond price'. What is more 'beyond price' than 'free'? Yet when you're offered something for free, you often spend much time asking 'why?'. Dan Ariely explores 'free' beautifully in a chapter of his book, Predictably Irrational. The chapter title is 'The Cost Of Zero Cost'. Taking something offered for 'free' is emotionally intense. 'Free' removes an entire host of commercial defence mechanisms from your armoury. 'Free' restructures the principal/agent relationship so now you wonder if you have an obligation to discharge. Zero cost is an asymptotic negotiating point where we become irrational. Marketeers are highly attuned to 'free' in many special offers. Ariely describes experiments from chocolate to book vouchers all showing that under 'free' conditions we make foolish choices. For example, when museums are free the queues are lengthy, despite the fact that museum fees may not be high. When confronted with 'free', we underestimate the value of our own time and accept much higher crowding.

Understanding another party's self-interest is crucial. We implore children to stop and think about why someone would offer something for free and tell them not to take free sweeties from strangers. One of the key pieces of information discovery in all negotiations is, 'what's in it for the other party?', 'what do they really want?'. The concept of self-interest and enlightened self-interest, looking to the longer-term, is fundamental to all of economics since Adam Smith. Enlightened self-interest is also crucial to public choice theory which recognises that problems with the behaviour of politicians and government officials are amenable to traditional economic analysis as they are self-interested agents and interact with social and economic systems under alternative constitutional rules.

The desire to establish self-interest is at its most paradoxical when politicians, businesspeople and non-governmental organisations (NGOs) meet. I've attended quite a few conferences where the first words to many presentations are, 'I work for the public sector' or 'I'm a public servant', or 'I represent an NGO', or 'I work at a not-for-profit'. These anti-commercial opening litanies are perceived to be important in allaying concerns about self-interest, and are then properly followed by what the presenter and their organisation do. Of course, despite the reality that I, like many other people, have multiple governmental, non-governmental and commercial roles, I often can't resist having a bit of fun at these conferences. I've been known to open some of my controversial presentations with, 'Unlike some of the previous speakers here today, I direct a for-profit organisation and can prove I add value to society?'. Once people establish commercial motivation, in many ways they can relax - the self-interest is clear. I can trust you to seek to make money.

[SLIDE: TRUST EQUATION]

It's worth dwelling for a moment on what it means to have multiple rounds. If we wanted to put forward a multiple round equation for reputation, it might be that 'reputation is the sum of all experiences less anticipation'. On the slide here, that is represented by the Sigma or sum of all experiences minus anticipations. Those mathematically-inclined amongst you can see all sorts of possible derivatives starting with variances and standard deviation.

A formal trust equation might be that trust equals obligation over reputation. If the obligation is small compared with the reputation, little trust is involved. When the obligation is large compared with the reputation, then much trust is involved. Note that an excessively large reputation contrasted with a tiny obligation implies low trust. However, each side of the equation has a principal and an agent. A poor down-and-out person trusting a very reputable wealthy person may be very trusting, while the wealthy person receiving the obligation may consider it a trifle. So the equation has to take account of relative trust, to whom. Personally, while I agree with the thrust of the equations, little things matter too. How much more reputation is gained when you realise that Megacorp will fulfil a small contract to the letter.

Despite a lot of promotion of the importance of high trust levels or good corporate social responsibility or good governance and possible relationships to price premia or share premia, the overall meta-analysis is still equivocal. [Some new techniques incorporating the importance of perception, for example factor assessment models as used by my firm Z/Yen for the Global Financial Centres Index, hold out hope of introducing some metrics in areas such as global financial trust or regulatory trust by differentiating between perceived trust and what instrumental factors would have predicted - see also Landon & Smith, 1998, and Quagrainie et al, 2003 below.]

Maister, Green & Galford [2001] put forward a 'trust equation' for professionals. First, they point out that in the earliest encounters you may place a great weighting on 'credibility'; the right diplomas, references or appearances can all improve credibility. Second, factors of 'intimacy' or empathy affect the weighting of experiences minus anticipation, for example you may be more forgiving of a family member failing to exceed anticipations, or less. Third, 'self-orientation' is crucial, closely related to 'self-interest'. A fourth element, of course, is 'reliability'. It's here that multiple rounds become intriguing, because multiple rounds are a game.

Trust In Games

[SLIDE: TRUST IN A GAME]

Game theory is a interdisciplinary branch of applied mathematics that is used by economists, biologists, engineers, political scientists, computer scientists and philosophers. Game theory uses mathematics, for example logic or probability, to model strategic situations in which an individual's success at decision-making depends on the decisions of others. Game theory started with zero-sum games such as chess, but encompasses non-zero-sum games, where cooperation increases the winnings of all players. Though games and theories about games go back millennia, the 1944 publication by John von Neumann and Oskar Morgenstern of Theory of Games and Economic Behavior gave the field a focus. Since 1944, eight Nobel prizes have been awarded to game theorists.

There are particular points in many games called equilibria. At an equilibrium, each player follows a strategy that they are unlikely to change. John Nash, one of the Nobel prize winners and the subject of a 1998 book by Sylvia Nasar made into the 2001 film A Beautiful Mindstarring Russell Crowe, identified game theory solutions now known as Nash equilibria. These equilibria are solutions where two or more players know the equilibrium strategies of the other players, and no player can benefit by changing their strategy while the other players remain unmoved. From Wikipedia, 'Amy and Bill are in Nash equilibrium if Amy is making the best decision she can, taking into account Bill's decision, and Bill is making the best decision he can, taking into account Amy's decision.' [http://en.wikipedia.org/wiki/Nash_equilibrium]. But a Nash equilibrium does not mean that the best cumulative payoff has been achieved. A classic example might be businesspeople competing under traditional economic assumptions. They could move beyond a Nash equilibrium and alter the rules of the game by forming a cartel. Then they would achieve a higher payoff.

A Mexican standoff or deadlock occurs when powerful forces oppose each other into a state of inaction. In a real-life business case I advised a large retail firm overseas competing largely against one other similar retail firm in the same country. Both had a few hundred stores spread throughout a landscape with few large cities but a large number of small rural towns. One firm or the other would often dominate a town. Unfortunately for both firms, there were a number of rural towns which could support between one and two stores. If one firm allowed the other firm to have a store without competition, then the competitor would gain a lot of money. Yet if both firms opened stores, both would lose money. Both firms had see-sawed back and forth over decades.

Of course, in the towns with two stores, people saved money until, as frequently occurred, both stores would close through lack of profitability, and then people would have to travel significant distances to shop when before they'd had a choice of two stores. In towns where one of the retail firms dominated, people paid more than they would have in two-store towns or the cities. Over time, without direct collusion, both retail firms recognised that two-store-town competition was a loser's game. Bar the odd two-store-town skirmish, things were static, a Nash equilibrium. To change the situation would require altering the rules of the competition, perhaps townspeople arranging for profit-sharing on exclusive stores (possibly good) or the two firms agreeing to adjust store locations to equalise profits and increase joint profitability (probably bad).

[SLIDE: MEXICAN STANDOFF]

The retail firms' situation reminds me of a 1983 film, WarGames, starring Matthew Broderick. In the film, Joshua, the computer character, concludes that when looking at thermonuclear war as a game, 'The only winning move is not to play'. Ultimately, call centres, deregulation and the internet did change the rules of the retail game and break the Nash equilibrium.

While game theory is a crucial analytical tool, there are numerous reasons why it doesn't work perfectly in practice. One of the key problems with analytical models of Nash equilibria is that the paradigm is one of known machines, e.g. Turing machines. Evolutionary systems have mutations that can change the nature of the game. Humans in commercial systems anticipate their situation. They can use their intelligence to go outside the system and mutate their roles or change the rules in the environment to their advantage. The 'feed through' of human systems means that feed-forward, i.e. anticipation, can be more important than feed-back. The intelligent ability to alter a system for personal benefit lowers the probability of Nash equilibria persisting over time in most human systems.

My mother and I used to have problems of trust when I was about five years old, a slightly earlier age than the son in my opening joke. I remember her wondering if she could trust me to keep my hands to myself in a shop or if she could trust me to stay in the car outside. This concept of 'trust' baffled me. Surely if she knew me well she could trust me to touch everything in the shop or to leave the automobile the moment she was out of sight?

Of course, we can trust someone to cheat, or steal or lie; or as in the joke about the wise mother, trust them to conceal which in turn forces them to reveal. The idea of trusting someone to be true to their nature, which in turn leads them to put unwarranted confidence in others, is the basis of many con-artist swindles. People who seek fame or money or status are frequently taken advantage of by con-artists 'working an angle'. Of course, these same verities of human nature feature in literature, television and film. The 1973 movie, The Sting, starring Robert Redford and Paul Newman, or the ending of the 1966 film The Good, The Bad And The Ugly, are good examples of top 100 films based around abuses of concepts of trust and deception, as well as numerous popular shows, such as Switch, Minder or Hustle, and the sub-plots of many police or detective shows.

Concepts of trust relate to many philosophical puzzles that range from Epimenides the Cretan's paradox of 'all Cretans are liars' through to Kurt Gödel's Incompleteness Theorem. A paraphrase of Gödel's Incompleteness Theorem applied to trust might read, 'We can never find an all-encompassing axiomatic system of trust, without recourse to systems outside it.' And many outside systems have been examined such as trusted third parties or security. Philosophers such as Hobbes and Hume have debated whether ultimate enforcement of trust resides in the sovereign or contracts or force. Kantian trust derives from principles and self-respect, while Foucalt's views on trust emphasise suspicion, visibility and accountability. Onora O'Neill [Reith Lectures (2002) -http://www.bbc.co.uk/radio4/reith2002/lectures.shtml] explored the contradictions in transparency, for example how accountability can lead to mistrust.

In summary, Bailey [2002] says, 'Recognising that human beings may take responsibility for how their behaviour influences others' decisions, however, offers us a way of explaining how trust can be rational. It thus also offers us a way of beginning to understand how trust can be genuinely cultivated and maintained.'

Trust In Volatility

[SLIDE: EXPERIENCES]

While trust and commercial relationships have complexities that mean Nash equilibria, though illuminating, rarely endure, simple contrasts of experiences and anticipations can shed more light on volatile commercial relationships and trust. Take this example representation of a likely range of experiences, a bell curve. You ask for car repair and the service is likely to exhibit a fairly classic statistical outcome.

[SLIDE: ANTICIPATIONS]

Now take this example of possible anticipation for a strong brand. You're pretty sure you know the brand well and what you will experience. The range is rather tight.

[SLIDE: GREEN SHOOTS OF ANTICIPATIONS]

However, reputation is about anticipation minus the experience. If your anticipation is high, then you're always going to be disappointed. If your anticipation is low, then you're likely to be surprised pleasantly. Clearly I'm ignoring some complexities such as empathy, e.g. you understanding that your car repair is delayed because the mechanic had to go to hospital.

[SLIDE: EXPERIENCES TIGHTER THAN ANTICIPATIONS]

Even well-informed people who have a good idea of the experiences they've had and what they can reasonably anticipate may jar with reality. This diagram shows the likely summary of people's assessment of reputation when experiences are 'tighter' than what was anticipated. Perhaps this is a company delivering a much more reliable and consistent service than most others. As you look at the statements, you can see that among the 'reliable', 'disastrous' and 'amazing' comments, there is a large zone of not-so-brilliant experiences with 'adequate - better than expected, though not expecting much' as well as reasonable experiences assessed as 'mediocre - had expected better'. In marketing terms this might be called 'brand dissonance'. Despite a reliable, consistent product, brand anticipations are not tight enough.

[SLIDE: ANTICIPATIONS TIGHTER THAN EXPERIENCES]

Now contrast brand dissonance with a situation where people's anticipations are 'tighter' than their experiences. In this situation, the number of reputation-enhancing experiences is much greater. You can understand why marketing people often go to great efforts not to raise anticipations too high. A fast-food restaurant advertisement shows it like it is - no tablecloths, no candles, no waiter service. And this leads us to volatility. If you consider the idea that 'tighter' equals lower volatility and 'looser' equals higher volatility, then you begin to see some of the reputational problems over the long-term. Sadly, the implication of the last two slides is that spending on tighter marketing initially has better returns than spending on tighter service delivery. There is clearly an interesting marketing paradox. Reputation risk management aims to lower reputational volatility to increase the odds of happier customers, but low reputational volatility conflicts with high-profile marketing founded on controversy designed to attract new customers.

[SLIDE: ACCUMULATED DIFFERENCES]

This slide summarises the importance of anticipation versus experience when evaluating trust. People ascribe amazing, adequate, poor, disastrous or reliable reputations based on combinations not of the outcomes, but of what they expected versus the outcomes.

There is an ancient Zen Koan, or puzzle that may enlighten you. 'When Banzan was walking through a market he overheard a conversation between a butcher and his customer. 'Give me the best piece of meat you have,' said the customer. 'Everything in my shop is the best,' replied the butcher. 'You cannot find here any piece of meat that is not the best.' At these words Banzan became enlightened.' [Reps and Sanzeki, 1998, Koan 31 - 'Everything Is Best', page 50]

The important observation is that a principal sometimes has to surrender their anticipations. By submitting to immense trust, and realising that a professional is trying to do their best all the time, a principal can suspend normal commercial monitoring with its accompanying overheads in time and costs. It may even be Zen enlightenment. On the other hand, to an outside observer, this can look as if the principal has a blind faith in market competition or professional service, i.e. blind trust. For those less philosophically inclined, perhaps you'll remember the words Richard Dreyfuss says to the liquor store salesman in the 1977 film The Goodbye Girl, 'A bottle of your finest cheap Chianti, please!'

[SLIDE: A PROBLEM CREATED - IS A MARGIN IMPROVED]

Technical Assistance Research Programs, Inc. [http://www.tarp.com/history.html] is frequently mentioned for a large number of customer satisfaction studies over the past three decades in the USA. Headline numbers often read that, on average:

a satisfied customer tells three other people they're satisfied;
a dissatisfied customer tells nine to ten other people - 13% tell 20 people or more;
a satisfied complainer will tell 5 people they're satisfied.

The implication of course is to trap complainers before they cause damage. This can be tough, as the same numbers point out that only 4% of dissatisfied customers complain - even those seriously wronged. In a different, but equally interesting case, one large hotel chain researched customer intentions after a visit to one of their hotels. They found that:

for customers with no problems during their stay, 89% intended to use the chain again;
for customers with a problem that had not been corrected, only 69% intended to use the chain again;
for customers who experienced a problem that was resolved before they left the hotel, 94% intended to use the chain again.

The opportunity behind these numbers is interesting, and the temptation too hard to resist. You can take advantage of volatility in anticipation versus experience. If you genuinely have a low volatility, reliable service, you should create problems. Everybody should have a problem. And there are companies that do create problems they know they'll solve. By knowing that 100% of their customers have a problem, they focus on getting the complaint rate from 4% or something, to more like 96%, and the problem resolution rate even higher. That's not hard if you've created the problem in the first place.

I've put in front of you the numbers for a hotel ensuring that everyone has a problem. Case 1, the top one, guarantees a problem, however small, for every guest. In Case 1, the hotel can focus on catching all problems and solving them. Ideally, the hotel generates a set of regular minor problems, perhaps even unnoticed, that will always be 'caught'. For example, the maid appears at the door with flowers a few moments after checking in - 'oh my, did we overlook putting fresh flowers in your room on time''. Having delivered the flowers, the maid returns a few moments later with complimentary chocolates to 'apologise' for the hotel's mistake.

Case 2, the lower one, deals with a more traditional hotel approach. As you may notice, in Case 1 returning customers are nearly 5% up from Case 2 and recommendations are 70% up. For those sensitivity trained among you, the catch rate in Case 1 can fall to 83% and remain a more positive strategy than Case 2. Of course Case 1 doesn't work so well if other service delivery is poor - multiple problems tend to be weighted more heavily. But then nor does Case 2. Further, not all problems are equal, so return and recommendation rates will vary. Finally, all this needs to be done with a straight face. Yet keeping the system hidden from the staff can affect morale - imagine a maid's opinion of a hotel that always asks her to deliver flowers just after guests arrive.

But this is not imaginary. There are such firms. If you order garden equipment via the internet from one online gardening shop, the online gardening shop's call centre rings afterwards to tell you that sadly they've mispacked your order. By mistake they have included a tool you didn't order which you are welcome to keep with their compliments. Meanwhile, they're sending along now the one bit of the order they missed, gratisnaturally, along with some free garden shears by way of compensation for your inconvenience. Now you move significantly up the scale of likely to re-order and likely to recommend. Moreover, many people probably tell their friends how they really gave the online garden store a hard time and through superior negotiating skills got free equipment.

I'll move on to one final observation before we move to conclusions, the hysteresis of trust. Hysteresis is a term that describes systems with memory. There are physical systems, say a pendulum clock, where the history of the clock over the ages has little bearing on today. By knowing some key variables about the pendulum you can predict the future state of the clock without having to know its history. On the other hand, systems such as tape recording rely on the history of the system being crucial to its future interpretation.

[SLIDE: BUILDING TRUST - SLOW PATH]

Hysteresis affects the way that trust is built. Similar sets of experiences in different orders can either build or destroy trust. Direction is important. Start anywhere on this diagram, say 'tough times' when both anticipations and experiences are dropping. These can be followed, rather obviously, by 'difficulties'. Yet from difficulties, as experiences improve, so can anticipations, leading to perceptions of 'commitment'. And from these perceptions of commitment from tough times emerges 'goodwill'. Trust is being formed on each cycle.

[SLIDE: SOWING BETRAYAL - FAST CROSSOVER]

Again, hysteresis says that the history of a system matters. Here we might start at the bottom with 'creating trust'. A series of experiences exceed anticipations leading to the principal 'trusting' the agent. However, a rapid drop in experiences leads to perceptions that the agent is 'taking advantage' of the principal. A further drop in experiences is accompanied by a fall in anticipations leading to the principal concluding that 'betrayal' is involved. The history of relationships, their sequencing and their perception, rather than their materiality, can either build or destroy trust, often exhibiting only minor differences between radically different interpretations. Hysteresis in trust seriously questions whether simplistic equations can ever capture the richness.

[SLIDE: VOLATILITY MAKES FRIENDS]

Just to bring this idea of the history of relationships making the difference between trust and betrayal, consider a relationship with almost no history. I ask you to remember the last time you made a new friend. How exciting he or she was. The crazy things they introduced you to, the new ways in which they looked at the world, books, films, music, art, jokes. The volatility of anticipation and the volatility of experience were both high - your new friend was exciting. Over time though, things change. [1] Constant surprise can get a bit wearing when you just wanted a quiet night's discussion and your friend is determined to blag their way embarrassingly into the latest hot nightclub or sneak up the back stairway to a church belfry. They've become dangerous. [2] Equally, you can get to know a friend well. Over time, as you know them better, they move from exciting to reliable. You can always count on them being up for - a film, or the opera, or to discuss a new book, or to get upset about the same politics you do. [3] Of course, you can also know a friend too well. Reliability can also slip into dull and boring. [4] But God forbid that they have a mid-life crisis with new friends or fast cars. Then clearly you don't know them any longer; they've become slightly dangerous; they seem to have betrayed your trust.

Trust In Money

[SLIDE: TRUSTING SOULS]

There are many complications to simplistic views of trust as an equation or hysteresis. Trends can go up or down in many situations. Frequently trust is gained slowly, but lost quickly, e.g. banking crises or share price falls or customer satisfaction. Just to take a simple example, I was once in a swish restaurant. On the way out I was unjustly accosted for not paying. It started as a simple misunderstanding by a member of staff but, instead of taking my word, everything had to be checked - somewhat spoiling a good night out and ensuring that restaurant would never be visited again. Why should I trust them if they don't trust me?

Likewise I ski a lot. The most amazing ski shops are the ones that take a chance - 'sure, try out our skis and don't worry about paying till you're done'. By increasing the risk in the trust equation to their detriment they gain a lot. A lot of businesses talk relationships, but act by transactions. A good example is a long relationship with a financial institution, say a business with a bank, which in the current crisis is frequently being rewritten with changes to covenants and terms. One friend of mine is responsible for financing a UK property company. His firm is well within its covenants based on historic valuations. However, after many years the bank wants him to either obtain new valuations at great cost, which they clearly hope are lower and break the covenants, or just accept the bank's new, much more expensive terms for finance.

This leads to my Zen koan - 'If you have some trust, I shall give you trust. If you have no trust, I shall take it away from you.'

Trust and money are uncomfortable partners. Ariely [2008, pages 67-68] amusingly describes an imaginary trust-busting scene, offering your mother-in-law three hundred dollars, no four hundred dollars, in appreciation of a wonderful family holiday meal. He points out that social and market norms conflict. We live in two worlds simultaneously, the social world and the market world. A simple solution is to keep the two separate. But this is often not possible. Today's financial crisis is, at heart, a crisis of trust. A couple of years ago, banks knew how badly they managed their own risks, how aggressively they'd priced their own assets, and how much their own bonuses depended on these aggressive valuations. 'If we're acting so irresponsibly, think how much more irresponsibly other banks must be acting.' Thus when repricing started, projections of social norms indicated that the repricing and credit terms should be more negative than market norms indicated. When banks failed to trust each other they withdrew credit, thus creating liquidity and solvency problems at fellow banks in a domino-effect.

Money and trust have strong parallels. Money is a unit of account, a medium of exchange and a store of value. Likewise, trust is a means of accounting, how we transact business over space and how we store value through time. As you might expect, just as technology has affected money - turning it from gold coins into bits and bytes - so too technology has affected trust. After sex and gambling, online technology makes its money out of trust. So-called 'trust technologies' include public key encryption, authentication and certificates. These are all mechanisms designed to automate and extend trust. New and interesting trust technologies inject human interaction and the analysis of human interaction into online exchange.

As a unit of account, trust accounting is increasingly being created by technologists - there are reputational ranking systems from point scores on Amazon, to supplier rating on eBay to collaborative filtering on many sites, to 'I hate' websites, to social networks with referral or testimonial systems, to reputational currencies under discussion, such as the Whuffie. As a means of transacting business over space, never before has there been a time when it's been easier to start a distant geographical relationship. With a credible website and reasonable links, you're prepared to learn about companies half a world away and entertain the idea of conducting commerce with them. Society is changing when you can find yourself trusting more people with whom you have no experience, e.g. on Ebay or Facebook, than your first encounter with your local policemen.

When it comes to reputation through time, customer appreciation and loyalty may become more persistent. We could move well beyond loyalty cards for cups of coffee or haircuts and the increasingly monetised transactions of air-miles to long-term value adhering to valuable customers. How much more valuable is a potential customer known to provide referrals for good service or products. Now imagine that potential customer can prove that they have a history of supporting good services and products, but demand a better deal in return. Indirect reciprocity on today's web, where ratings are provided honestly in order to help strangers, may alter over time. Raters might get direct benefits, or be able to use their selflessness to increase their own reputations. The increasing durability of online data begins to hint at the idea that ruptured relationships may be harder to gloss over in future. I mentioned recently to a friend the rising ethical issues for young people on many social networking sites. These young folk may be unable to deny in the future a relationship they had with a fraudster or axe murderer or terrorist today.

Likewise technology challenges governments, organisations and professions. Numerous surveys (e.g. Edelman's Trust Barometer) show decreasing levels of trust in institutions and professions ranging from clergy to politicians to businessmen to journalists. This is not surprising. Lower cost and more prevalent information permit people to identify more easily those who break trust over time and space. We can find information about one killer doctor that may have been hard to find in times gone by. We have yet to adapt to the idea that that information may not change the odds of encountering a killer doctor.

[SLIDE: VIX TO REALITY]

The relationship between trust and money also turns up in confidence or suspicion over the future value of money. VIX is the ticker symbol for the Chicago Board Options Exchange Volatility Index, also called the 'fear index'. The VIX is calculated from the implied volatility of S&P 500 index options and reflects expected future volatility. A high value corresponds to an expected more volatile market, implying more costly options. Likewise, inflation uncertainty, as measured here by the standard deviation of the previous 12 months inflation, correlates reasonably well with VIX. There are many other possible indicators, base rates, inflation indices, yield curves or forward prices to name a few.

At one point in the very early 1970s I remember my father explaining that a combination of tax deductible mortgage interest payments and inflation rates rising faster than interest rates meant that government and lenders combined were paying him to buy his house - 'the late 1960s and 1970s gave debtors a free lunch as the real value of their debts and interest payments declined. While American home purchasers in the mid seventies anticipated an inflation rate of at least 12 per cent by 1980, mortgage lenders were offering thirty-year fixed-rate loans at 9 per cent or less.' [Ferguson, 2008, page 253]. My father and others rather naturally took advantage of these rates. Inflation became a sign of mistrust which, when governments turned to controlling stagflation, led to great property and savings & loan collapses.

We have moved from trust being between individuals, to spanning society. At times of low confidence and high uncertainty, the VIX and inflation expectations tend to rise, societal trust falls. With more time we could explore trust within an organisation, as Covey & Merrill [2006] and many other organisational anthropologists do, distinguishing high trust from low trust organisations, but I'd like to turn to some concluding thoughts.

Trust Me, I'm Human

[SLIDE: TRUST ME, I'M HUMAN]

Do we want a high trust society or a low trust society? This question has been tackled by many commentators. The general consensus is that a high trust society is a preferable place to live and probably more efficient and effective. However that does not mean that trust is an unmitigated good, nor that we shouldn't retain some mistrust. If you could imagine a 'no fraud' Britain, you could equally imagine a national of gullible folk. In a 'no fraud' Britain, you can imagine corrupt Swedes and Swiss leaving restaurants without paying. Fraud is a virus that in many ways helps the host to keep up its defences. What we want is rising levels of trust without losing basic levels of suspicion.

In a high trust society there is not enough incentive for competitive checking, not enough suspicion, too much confidence. In a high trust society, competition may not work effectively in keeping self-interest in check. In a low trust society, constant checking wastes resources, whether it's because people are being ripped off by one another, or the government has created a low trust society by taking resources through inflation, thus resulting in wasteful anti-inflationary planning.

Commerce is about people and trust, not just money. Low trust and high value transactions are almost wholly consigned to the criminal backwaters of society. Equally, we will always have large numbers of low trust and low value transactions, ranging from street stalls to chance encounters. When society functions well we have large numbers of high trust and low value transactions being conducted efficiently, e.g. newspapers, prescriptions. When times are good we have quite a few high trust and high value transactions, e.g. long-term construction contracts, that are precisely the transactions to suffer most when society as a whole is leaking trust.

Tonight we've had a whirlwind tour of the complexity of something basic to commerce - trust. We've seen that transactions over time, volatility, games and money all complicate things. At the end of the day, we must decide whether commerce corrodes or enhances trust. We must decide between the Irish dramatist Oliver Goldsmith's (1730-1774) assertion that 'Honor sinks where commerce long prevails' [The Traveller,http://1picture.org/poems/golds02.html] or another Irish dramatist, George Bernard Shaw (1856-1950) who wrote, 'He's a man. I know him: his principles are thoroughly commercial.' [Man And Superman, Act 4].

For me, commercial self-interest ensures that successful commercial people must maintain a basic level of trustworthiness, while competition ensures that basic trust levels increase over time. Or am I just too trusting?

Thank you.

Discussion

How can we know when we're too suspicious - the opposite of 'too trusting'?

Can we ultimately automate trust?

Further Reading

ARIELY, Dan, Predictably Irrational: The Hidden Forces That Shape Our Decisions, Harper Collins (2008).

COVEY, Stephen M R and MERRILL, Rebecca R, The SPEED of Trust: The One Thing that Changes Everything, New York, Free Press (2006).

FERGUSON, Niall, The Ascent Of Money: A Financial History Of The World, Allen Lane (2008).

LANDON, Stuart and SMITH, Constance E, 'Quality Expectations, Reputation, and Price', Southern Economic Journal, Volume 64, Number 3 (January 1998), pages 628-647 -http://www.plu.edu/~reimanma/doc/quality-expectations.pdf

MAISTER, David, GREEN, Charles and GALFORD, Robert, The Trusted Advisor, Simon & Schuster (2001), ISBN 978-0743212342.

MARKOSE, Sheri M, 'Gödelian Foundations of Non-Computability and Heterogeneity In Economic Forecasting and Strategic Innovation' - Paper Presented at Computability in Europe 2006: Logical Approaches to Computational Barriers (2006) - http://www.essex.ac.uk/CCFEA/Teaching/CF902/2006-07/Spring%20Sheri%202007/MarkoseGodel%20paper.doc

QUAGRAINIE, Kwamena K, MCCLUSKEY, Jill J and LOUREIRO, Maria L, 'A Latent Structure Approach To Measuring Reputation', Southern Economic Journal, Volume 69, Number 4, Southern Economic Association (April 2003), pages 966-977 -http://www.ses.wsu.edu/PDFFiles/JournalArticle/McCluskey/measuring%20reputation.pdf

REPS, Paul, and SENZAKI, Nyogen, Zen Flesh, Zen Bones, Tuttle Publishing (1998).

WARD, Aidan and SMITH, John, Trust and Mistrust: Radical Risk Strategies in Business Relationships, John Wiley & Sons (2003).

Further Surfing

BAILEY, Tom, 'On Trust And Philosophy', BBC and Open University (2002) - page 9, - http://www.open2.net/trust/downloads/docs/ontrust.pdf

BRAYNOV S and SANDHOLM T, 'Contracting with Uncertain Level of Trust', Proceedings of the first ACM conference on Electronic commerce (3-5 November 1999) - in Robert Ghanea-Hercock -http://www.santafe.edu/network/hercock.php

Edelman Trust Barometer 2009 - http://www.edelman.co.uk/files/trust-barometer-2009.pdf

More on 'Trust Equation' -http://www.nationalunderwriter.com/pandc/Awards/AgencyBestPractices.pdf

Reputation Management -http://en.wikipedia.org/wiki/Reputation_management

SEAMAN, Paul, 'Would You Trust A Trust Suvey?' (28 January 2009) -http://paulseaman.eu/2009/01/would-you-trust-a-trust-survey/

Trust in general - http://changingminds.org/explanations/trust/trust.htm

Thanks

It would be interesting to see if they trusted me to remember to thank them for invaluable assistance - Ian Harris,

Paul Hirst, Bill Joseph, Nicholas Mainelli, Sheri Markose, Rupert Stubbs, Aidan Ward - whereas others of you may be right, I forgot.

©Professor Michael Mainelli, Gresham College, 2 March 2009