Post No.: 0816
In a joint evaluation when comparing two scenarios – one where an innocent person gets shot after stumbling into a robbery at a store where that person frequents, and one where that innocent person gets shot after stumbling into a robbery at another store where that person had to go to because their regular store was closed for renovations – most people will accept that it’s irrelevant which store that person sustained a crippling injury in when it comes to deciding the amount of victim compensation.
Yet in a single evaluation – where two separate groups of people are asked to evaluate only one or the other incident, not both – they’ll tend to overall award a much larger sum to an innocent person who sustained their injury in a store they rarely, compared to regularly, visited. Hmm.
The difference is because situations that require making explicit comparisons require the slow thinking of ‘system two’, and carrying out a joint evaluation of the two scenarios gives us a chance to examine our moral principles and work out that the location where someone sustained an injury should be irrelevant when it comes to deciding on the amount of victim compensation.
In the single evaluation, poignancy (which is closely related to regret) is at play due to the thought of ‘if only they shopped at their regular store’. The substitution heuristic and intensity matching then translate the strength of one’s emotional reaction onto a monetary scale. (The emotional appeal of pandas is also why most people are more willing to contribute to their protection than to bees, even though bees are more important to us.)
Almost everyone making a joint evaluation would say that poignancy should be an irrelevant factor. Unfortunately, this only becomes apparent if the two scenarios are seen side-by-side – and this isn’t how real life usually works! We normally experience life in a ‘consider one thing at a time’ way, in which contrasting alternatives that might change our mind are absent due to ‘what you see is all there is’ – which in turn can result in us making internally inconsistent decisions (that we won’t be aware are inconsistent!) So the beliefs we endorse when we reflect about morality don’t necessarily govern our present emotionally-led reactions; and the moral intuitions that come to mind in different situations aren’t always internally consistent.
This discrepancy between single and joint evaluations can result in reversals of judgements and preferences when people switch between making the two types of evaluation e.g. most people would prefer to take a smaller but safer bet to a larger but riskier bet when asked to make a joint evaluation; yet if they were made to imagine owning only one of those bets and putting that up for a trade, most people would set a higher value on the second bet than the first! This is perhaps because the size of the potential prize provides an anchor, hence the larger but riskier bet will be valued higher.
When evaluating cases in isolation, we’re not aware of the anchoring effect from the single number we’re exposed to. Plus it’s only in joint evaluations can we notice that the second bet is relatively riskier too. Factors, like anchors and emotional reactions that sway our ‘system one’ – that can cause a discrepancy between our judgements in single evaluations – are suppressed or deemed irrelevant when the options are evaluated jointly, when a more effortful and careful analysis involving system two will be employed.
So separately-made choices aren’t always consistent or coherent – they depend heavily upon the context in which they’re made (contrary to classical economic theory). Perception is always about relative comparisons, not absolute states, hence what’s considered ‘valuable’ can only be achieved by comparing something relative to something else that one has currently brought to mind. And the more specific the alternative, like when making a joint evaluation between explicit options, the easier it’ll be to make a consistent relative valuation.
But in real life, most decisions are made in single evaluations, not joint evaluations where every opportunity cost is made explicit to us. This would prohibitively take too much time and effort to conduct every time anyway. Moreover, in many real-world scenarios – how can one ever truly know if one has considered every single possible alternative in existence before making the best choice?!
Our system one automatically breaks things into categories, and each category will have its relevant norms according to our own perceptions (stereotypes). Something being ‘large or small’ or ‘good or bad’ will depend on what context, evoked category, and associated norm, we’re talking about e.g. ‘1km wide’ will be small for a border, but large for a river.
In single evaluations, we’d say a 6-year old who is 5’ is tall, and a 16-year old who is 5’1” is short, because they’re compared to different norms. But if asked if the first person is taller than the second in a side-by-side comparison, we would say no. So the process by which objects/events recruit their context of comparison can lead to incoherent judgements – judgements and preferences are coherent within categories but are potentially incoherent when the things being evaluated belong to different categories.
When asked to make a choice between two desserts, we’ll find it easy to make a consistent choice because the items belong to the same category – there’ll be no preference switch because different desserts are compared to the same norm and implicitly compared to each other in single, as well as joint, evaluations. But if asked to make a choice between a dessert and a main course, our decision will be less stable because they’re not natural substitutes – they don’t belong to the same category hence the potential for a preference switch if asked in different contexts.
If asked to make a donation to a ‘save the endangered bilbies’ fund, we’ll compare this to similar requests in the ‘nature charities’ category. And if we rate furry bilbies highly compared to these other endangered animals and environmental causes, we’ll use substitution and intensity matching to convert our emotional intensity into a monetary amount and donate what we’ll consider as a ‘decent’ amount (which may differ from what we consider ‘decent’ if we were to make donations to a political party or local food bank, for instance).
But if asked to make a donation to a ‘save farm workers from skin cancer’ fund, the category this request belongs to may be the ‘public health’ category, and when compared to other issues in this category, it may seem like a very low priority (lower in its respective category scale of public health issues than bilbies in its respective category scale of nature charities) hence, via intensity matching, we may offer an objectively smaller contribution to this fund than to the bilby fund.
So in narrowly-framed single evaluations, bilbies will elicit a higher intensity matching score and so will tend to attract much larger donations than for farm workers. Yet in a more broadly-framed joint evaluation, most people would rather donate more to farm workers than bilbies because farm workers are humans whereas bilbies aren’t! The comparison didn’t seem relevant or salient, and wasn’t made apparent, until we were asked to make a joint evaluation between bilbies and farm workers. Joint-framing changes the representation of the issues – allowing us to see a bigger picture and put our emotions and preferences into a broader context.
In single evaluations, if asked to value a ‘like new’ condition dictionary from 2015 with 50,000 entries – most people will value it higher than a ‘cover torn, otherwise like new’ condition dictionary from 2015 with 100,000 entries. Yet they might prefer the latter in a joint evaluation. Standalone information, like the number of entries here, is given no weight in single evaluations because ‘little or lots’ can only ever be determined relative to something else – and only in a joint comparison can we see that the number of entries is more important to us than the condition of the cover.
If fixated narrowly on ‘buying a camera’, one might do comparisons in that product category and settle on buying a £800 SLR camera. Yet if one thinks a bit more broadly, one might think £800 could get a smartphone that’ll also have an adequate enough camera for one’s intended purposes too.
The problems of predictably inconsistent judgements are prevalent in legal cases – in mock juries, awards to victims of physical injury will increase when made in joint evaluation with victims of financial loss, compared to in separate single evaluations of each. Yet in real life, jurors are actually prohibited from considering other cases! We can only see this absurdity when viewing cases together instead of separately.
There may be internal coherency within institutions but inconsistencies can be apparent when comparing between the compensations offered by different institutions, in a manner reflected by politics and history more than any overarching concern for fairness.
So rationality, fairness and reasonability are better served when employing broader and more comprehensive frames, such as joint evaluations. But of course we must be cautious when someone who controls what we see and make comparisons with has a vested interest in what we choose e.g. retailers routinely manipulate the context in which customers assess products in order to influence our preferences and justify their prices – something may be cheap compared to other items in their own stores, but compared to other items in other stores it could be expensive.
Rational agents shouldn’t be swayed by how identical offers are phrased e.g. carob is carob, and it shouldn’t matter if someone calls it a locust bean because it wouldn’t affect the true state of the world. However, real humans are constantly swayed by such altered presentations of things, hence the massive marketing and PR industry – indeed if humans were more rational, this industry wouldn’t be so massive and marketing materials wouldn’t be so elaborate.
A rational agent wouldn’t care if a bottle of water was advertised against a tropical backdrop with sexy celebrities doing something hilarious and the product was called ‘the saviour of hydration’ – it’s just a bottle of water, and they wouldn’t pay more for it than the utility of quenching a thirst. Woof!
They wouldn’t pay more for a ‘diet’ product because most of the time it’s just a smaller portion of something. Or be influenced by the smell, lighting or music in a store. Or care if a brand sponsored a team or event, as if being associated with a ‘fast and winning’ car meant the products are ‘fast and winning’ too. Or be influenced by effects like the decoy effect (presenting an asymmetrically dominated choice in order to make a Pareto dominating choice appear more attractive) or the power of environmental nudging (which can be used to encourage or discourage behaviours e.g. supermarket tricks to get customers to purchase the most profitable products on the shelves), and more.
But real customers are persuaded by such tactics, and are most of the time unaware they are too. So much that we’re unaware of affects us unconsciously yet is so critical to our experiences, feelings and in turn decisions and actions; just like a film’s musical score. Our beliefs aren’t always reality-bound. The benefits we’re paying for may be only in our minds, and were planted there via framing effects, peripheral cues and other irrelevant influences.
Logically equivalent statements can evoke different reactions and associative memories in system one e.g. one would say, regarding a cricket match, if ‘team A won’ then it’d mean exactly the same as ‘team B lost’ – but these different framings actually evoke different associations, of ‘winning’ and ‘losing’-related emotions and memories respectively. Therefore – irrationally – logically equivalent statements can evoke different meanings and reactions in us. Our preferences can therefore switch even if an offer is logically identical, with the only difference being its presentation or framing; which is contrary to rational-agent theory.