[ Jocelyn Ireson-Paine's Home Page | Free software | Publications | The Lives Behind the Numbers On the Screen: Illustrating the Social Consequences of Economic Change By Telling Stories On the Web ]
And yet... In all the houses and streets there is peace and quiet. Out of fifty thousand townsfolk there's not one ready to scream or protest aloud. We see people shopping for food in the market, eating by day, sleeping by night, talking their nonsense, marrying their wives, growing old, complacently dragging off their dead to the cemetary. But we have no eyes or ears for those who suffer. Life's real tragedies are enacted off stage. All is peace and quiet, the only protest comes from mute statistics: so many people driven mad, so many gallons of vodka drunk, so many children starved to death.
Oh yes, the need for such a system is obvious. Quite obviously, too, the happy man only feels happy because the unhappy man bears his burden in silence. And without that silence happiness would be impossible. It's collective hypnosis, this is. At the door of every contented, happy man there should be someone standing with a little hammer, someone to keep dinning into his head that unhappy people do exist - and that, happy though he may be, life will round on him sooner or later. Disaster will strike in the shape of sickness, poverty or bereavement. And no one will see him or hear him - just as he now has neither eyes nore ears for others. But there is no one with a hammer, and so the happy man lives happily away, while life's petty tribulations stir him gently, as the breeze stirs an aspen. And everything in the garden is lovely.Anton Chekhov, "Gooseberries"
What Chekhov proposes is something that I tried to do, and that is explained in The Lives Behind the Numbers On the Screen: Illustrating the Social Consequences of Economic Change By Telling Stories On the Web. What is on the page you are reading now is a questionnaire designed to elicit the economic knowledge I need in order to generate these stories.
This first section is concerned with classifying VE's outputs (i.e. variables such as unemployment and inflation) into economic regimes such as boom or recession. Then in the next section, I ask you to associate each regime with its own particular stories. Then to generate stories, I just need to code conditions that implement your classification, and then pick stories from the regime thus selected. (I do allow for more subtlety than this, but it's a quick way to get started.) The idea is that these stories will appear (quite likely after I have rewritten them for style, and possibly with some programming to fine-tune key words to the circumstances) in a newspaper front page generated from VE's output.
I also want words or phrases to describe values of the output variables, e.g. to classify unemployment as "high" or "low". I need these for text to slot into headline and story templates.
Please give the names and properties of all the macroeconomic variables in VE's output you want to take into account when generating the stories. In this and other tables, I've put examples to show the kind of thing I want. They are probably wrong!
|Variable||Symbol||Units||Min value||Max value||JND(*)|
|Unemployment||U||% of workforce||0||100||2|
(*) JND is the "just-noticeable difference". It's the smallest value D for which you would say that V+D or V-D is significantly different from V. E.g. it probably doesn't matter if unemployment is reported to have gone from 2% to 2.001%; it probably does matter if it goes from 2% to 3%. If the smallest change you'd say makes any difference is 0.1%, that's the JND.
Now I need a set of regime names, and of conditions which, given the relevant variables from a VE forecast, will tell me which regime it belongs to. You can think of this as a function which takes the variables defined just above and returns a string naming a regime.
|Boom||U < 2 and I < 2|
|Recession||U > 5 and G < 1.5|
It would be useful to have general headlines or other sentences appropriate for use in each regime.
|Boom||It's boomin' bootiful!|
|Recession||Prophets of doom forecast economic gloom.|
|It's throw-yourself-out-the-window time, folks!|
|Recession and on and on... and recession and on and on|
If there are words in a sentence that might change with the values of one or more of the variables, indicate this by writing a condition into the text.
Also, if you need, you can write specific conditions on the variables in the "regime" column, rather than regime names.
Now I need information that will allow me to generate words for values of individual variables in the output. We talk about unemployment as very low, low, normal, and so on; inflation as low, normal, high, hyperinflation and so on. For each variable, please think of words or phrases that would describe a range of values, and give the endpoints of this range. (Obviously, this is approximate, rather like describing temperatures as cold, cool, warm or hot, or people as short, tall, or very tall.) Give as many ranges as you can (preferably at least seven for each variable), and don't worry about using synonyms: it doesn't matter if the ranges overlap. I can then use these ranges to generate names for values, by searching all the intervals for the one in which the value is most central, then using the words attached to it.
Use the right units!
|Variable||Word or phrase||Min||Max|
It would be useful to have general headlines or other sentences appropriate for use in each interval. Here we are concerned with just one variable, e.g. inflation, per headline, rather than the economy as a whole.
|Unemployment||Very low||Chancellor gets full marks for full employment!|
|High||In the doledrums|
|A doleful Budget|
|Inflation||Very low||Chancellor deflates his critics - and the economy!|
|High||Pound falls through the hole in your pocket|
|Pound takes a pounding|
If there are words in a sentence that might change with the value of the variable, indicate this by writing a condition into the text.
Also, if you need, you can write specific conditions on the variables in the "interval" column, rather than phrases such as "very low".
Now I want to start linking VE's outputs to what people would be doing in the economy. I want to start by simply associating stories (at least some of which will be educational, I hope) with particular regimes. I've put a number of examples below, and would like you to fill in the table so it covers all the regimes. Use whatever style you want.
Some stories can be about the economy as a whole, but I also want to show how people's lives are affected, especially in their emotional reactions (see more about this here). The kind of response I want to get into at least some of the stories is exemplified by this quote (from Radio 4 comedian Mark Steel's book, "Reasons to be Cheerful"):
There are several episodes in the table's examples below like this.
In one sense being unemployed is the hardest job of all, in that it's relentless. It wouldn't be so bad if the hours were the same as other jobs, so you spent the day in a numb trance or walking to Camberwell to get cheap onions, but at five o'clock reverted to a normal human being with forty quid in the bank and a reason to wash. What a joy it would be to wake up on a Friday thinking 'one more day and then a whole weekend of having enough money for a pint before being back in the sleeping bag'. And at least you'd feel you mattered if a supervisor came round once an hour yelling at you to get back to sleep.
Instead, unemployment impregnates you with an infectious sense of worthlessness that spreads through your whole body to leave a numb vacant slouch, because you can spend all day in a sleeping bag and no-one's going to complain, or even notice.
|Boom||"We need massive new road-building programme", says transport minister as he admits Britain's creaking infrastructure can no longer cope.|
|Booming freight traffic pushes cyclists off road and into hedges: "Growth has its costs, and this is one we don't wish to pay", says head of CTC.|
|Fashion: Milan 2007 sees return of mini-skirt, outrageous reds and scarlet stripes in mens' Autumn collection.|
|Down with the urban dirge: new wave of optimism hits pop.|
|Mr Bruce called me into his office earlier today. "It's good news", he said. "From Monday, we'd like you to join Quality Control. We'll be giving you a pretty decent rise, and we think you'll find the job a good deal more interesting. The work will be challenging, but we have every confidence that you'll manage it". The economy was bubbling like yeasty dough. New jobs were being created at a rate of knots, office buildings sprouting like dandelions, cafe's hiring staff by the hour to feed the hungry hordes. Hungry hordes who were gulping down Bruce's Better Bakes by the bellyful. "We're desperate to attract more staff", he said, "but it takes time to find and train them, and there's so much competition. We'll soon have it sorted, but we hope you won't mind the odd busy day in the meantime."|
|There's a strike at the Sandy Lane depot - profits are rising, and they want their share. Delivery drivers who were turned away by the pickets swear they saw the leader standing on a table and yelling "Not a single biscuit shall leave Sandy Lane". So we were left to pick up the crumbs, as it were.|
|Delivery men stuck in traffic jams. Strikes. The people who provide us with freshly shelled eggs (safety regulations dictate you can't shell your own eggs) have a run on supplies and have had to cancel - men may work faster but the hens will not. Late nights. Cramming down a cold Pat's Perfect Pasty for lunch as I inspect the rotary moulds. It's not worth it.|
|I have asked for my old post back. The pay isn't so good, but at least I'll see my son before he's asleep each night. I look at the new QC supervisor and think "Do you never ever ever, as you fill in your hundredth Eccles Cake Inspection Docket of the day, when you wish it were only half-past five but know it's half-past nine, consider grabbing that filing cabinet, setting fire to the thing, and lobbing it straight through the window?". I know I did.|
|Hyperinflation||"Dear Editor, last night I saw the following sign in my local: ''Never mind the Campaign for Real Ale, what we need is a Campaign for Real Money!''".|
|Software company launches automated menu price update program and display for pubs, cafe's, reports LVA.|
|"Dear Editor, Last night, I noticed my eight-year old son building a pretend fort out of bundles of ten-pound notes. In Tuesday's business supplement you reported on Lego (TM)'s profit downturn. Could these be related?"|
|Recession||Sharp downturn in holidays - travel agents report record losses.|
|Sales of antidepressants reach all-time high.|
|A memo has come round from the management. The finances are in a bad shape, there are to be lay-offs. As dusk falls, I feel a gloomy foreboding air to the day that matches the pouring rain.|
|On the dole, you literally run out of money. I have had frustrating conversations about this, in which friends will shrug their shoulders and say it's a nuisance when you have no money. "But I don't mean no money until I get to the bank", I say, "I mean NO MONEY". In that situation, one of the most frustrating pieces of advice is to be told that X sports shop had a sale of trainers or Y record shop was doing a promotion on cut-price CDs. Because when you have NO MONEY, you go without. You might just as well announce that Swan Hunter have a sale on battleships: Iraq War seconds, only £80 million quid until the end of the month.|
|I have given up trains, buses, and beer; now I am giving up smiling, in preparation for all the other things I shall give up. Giving up getting up.|
|Velvet Vampires Dance in the Dark: Gothic renaissance amongst the economic gloom.|
If there are words in a sentence that might change with the value of a variable, indicate this by writing a condition into the text. This could be useful for inflation-dependent stories, where the frequency at which loss of real value becomes significant can be calculated from the inflation rate.
Also, if you need, you can write specific conditions on one or more variables, or use the names of the intervals (e.g. inflation "very low") in the condition column.
You may want to give different types of story for extreme conditions that VE can predict but that have rarely or never been experienced in the UK, from those for more typical conditions. It's all very well to illustrate hyperinflation with stories reverse-engineered from 1920s Berlin and 1950s Hungary, but what can one say when VE predicts economic circumstances very similar to today's? This is an educational decision. Perhaps the thing to do is to devote the "typical" stories to explaining points such as the difference between frictional, seasonal, and other kinds of unemployment, for example. (We also need to take care that the reader does not get the impression that if unemployment is up at 20%, say, no-one at all has a job.)
In the table above, there are two sequences of episodes that run on from one another, the one about the baker and the one about the unemployed person. If you feel up to it, I would like to have a few such sequences for different economic conditions. These can be included in the output as a kind of virtual diary reported in the paper, and they will give a sense of coherence, making the paper more than a set of unrelated events. Even more useful would be if you can indicate alternatives that might occur in such a history: e.g. a person might start suffering from depression as a result of traumatic economic events (there is a theory of depression that sees "life events" such as the loss of a job or the need to move house as triggers for depression).
Most people love spotting running jokes in comedy, whether in prose, films, or cartoons. They must do, otherwise comedians wouldn't spend time perfecting such jokes. This must be something to do with the brain's deep need to recognise patterns. Perhaps we can capitalise on the sense of satisfaction such pattern recognition brings, by showing how the same characters' lives vary under different circumstances and giving students the fun of spotting the relationship. So if you have given a story about cyclists being pushed off the road by heavy freight traffic when business is booming, it would be nice to have a story about the same cyclists under different circumstances, e.g. noticing how quiet the roads are during a recession. The game author Jim Gasperini has apparently used this technique in his game Hidden Agenda, a simulation of Central American politics where students try their hands at governing the simulated country Chimerica. Discussing the open-endedness of such simulations, Gasperini writes:
Even in the best 'interactive fiction', once all the puzzles have been solved the plot is revealed in all its naked linearity. A finished 'closed-ended' work is like a punctured balloon, emptied of all ambiguity. There is little reason for anyone to go through it again. By contrast, an 'open-ended' work becomes more ambiguous, not less, the more it is played. It is through repeated playings, comparing different plots chosen through the same web of potential plots, that the experience becomes most meaningful. This can be most clearly seen in the genre known as 'simulations'...
Each subsequent time the player enters the election campaign comparisons naturally arise between what happens this time and what happened other times. This serves to deepen the player's awareness of the range of structural possibilities.
You might want to think of the world under one economic regime as being in an alternative universe from the world under another: almost the same people, almost the same objects and events, but with small but significant differences. (Science fiction readers seem to love reading alternative universe stories and picking out these points, which again suggests its potential for motivation.)
Another variation would be to tell the same events from different points of view. For example, the stress experienced by an unemployed person might be recounted directly by them via a journalistic interview, or via the doctor treating him (perhaps he writes to the paper's agony aunt, for example), or via the local pharmacist kept open all hours dispensing anti-depressants and herbal anti-stress pills, or via a business report about drug company profits...
Perhaps students will enjoy pushing the economy to its limits to see how the simulated people suffer, in much the same way that children enjoy crashing toy trains. I am sure that this - breakability - is something that motivates in game-playing, and it may do here as well, providing another reason to include stories about dire circumstances.
To lighten the output, give a sense of fun, and add to the richness of the simulated world (or "explorability": like fun and breakability, a motivating factor in computer games), it would also be useful to have suggestions for fillers: stories with no direct economic content. Indicate if there is a context such as the weather from which one might generate alternative versions. One might have a common thread running through several fillers, as shown below.
|London taxi driver joins Foreign Legion "to escape congestion charge".|
|Floods demolish railway embankments: hundreds of trains cancelled.|
|Crowds cheer as floods wash away hostel for asylum seekers.|
|Fat-cat councillors escape dismal weather on "fact finding" trip to Hawai.|
|Fall of Roman Empire was "due to illegal immigrant barbarians", claims populist historian.|
|Coffee "is Britain's substitute for sunshine" claims Italian celebrity chef.|
|Record heatwave buckles rails: hundreds of trains cancelled.|
|Fat-cat councillors escape scorching heat on "fact finding" trip to Alps.|
|Dry summer and early Autumn bring record leaf-fall: hundreds of trains cancelled.|
|No room at the inn? "Mary forced to give birth in stable because inn full of asylum seekers", claims MP.|
Now I want to add more structure to the output. It will be useful for students to see how conditions are perceived by the main politicians, and how different politicians (and others) differ in their beliefs. So I want a cast of national figures. Pick or fill in at least three, preferably ones who differ at least slightly in their economic views.
We assume all these characters know (or think they know) something about the economy and when it's going well. So I also want some notion of how they do so: a "happiness index" which calculates their satisfaction with the economy from the output variables. This can be used to automate the production of some interesting headlines. At its crudest, a test that just distinguishes satisfaction from dissatisfaction would do: but you may prefer to grade more finely, for example to say that the TUC measure of satisfaction is related to the unemployment percentage.
|Chancellor||happy if I < 2; unhappy otherwise.|
|Prime Minister||happy if boom; unhappy otherwise|
|Head of CBI||...|
|Head of TUC||% happiness = 100% - U|
|Leader of Opposition||...|
|Head of ECB||...|
Some individuals' views might depend on forecasts for previous years too. For example, if the head of the ECB is interested in Britain's monetary stability, perhaps his happiness index might calculate it from the previous five years of VE's forecasts.
Now I want some advice about how to correct things when the economy goes wrong. It would be good to have advice for all the regimes deemed to be undesirable, and to have conflicting pieces of advice from different characters.
|Recession||Public spending to stimulate growth||Head of TUC|
|Lower taxes to stimulate growth||Head of CBI|
If there are words in a sentence that might change with the value of a variable, indicate this by writing a condition into the text.
If you need, you can write specific conditions on one or more variables, or use the names of the intervals (e.g. "inflation is very low") in the condition column.
The previous section uses a pretty simple approach to generating stories. In this one, I'm trying something more complicated which takes into account the "ontology" of the entities involved. By this, I mean the way they're classified as objects, processes, events, and so on. For example, we might say that stress can be one of the costs of unemployment, and traffic congestion can be one of the costs of inflation. However, these are different kinds of entity - at the micro level, unemployment is a change in a person's state, whereas inflation isn't - and for a robust extendable program, ought to be treated accordingly.
Let's start by thinking about costs. To start with, I just want something like the conventional list of costs taught in most textbooks:
|Unemployment||Loss of skills|
|Loss of output|
When filling these in, you may think there are also costs to, say, low inflation or low unemployment. If so, classify them separately under appropriate left-hand headings.
Although it's easy to write down a list of costs, it's not so easy to turn it into a program. Unemployment causing a person's skill to decrease because it decays in the absence of practice is a different kind of process from unemployment causing loss of output because of the aggregate effect of lots of ex-workers no longer being able to afford to buy beer or CDs. Traffic congestion is a change in a different kind of entity from loss of skill. And so on.
Let's start with the microeconomic costs, those which it does make sense to apply to individual people. Imagine someone (yourself!) facing each of the costs in the list above. At its most general, a cost means that something has changed in a way you dislike. Either some property that you deem desirable has decreased, or some property that you deem undesirable has increased. This property may be a property of you, or of an object or objects in your environment.
Let's begin with unemployment. This is a property which applies to individual people in a way that inflation and growth do not. We also have an aggregate unemployment count which is the macroeconomic variable. So we should distinguish them. I'll use a pseudo-object-oriented notation to do so, together with an if-then condition-action notation for rules about what happens as a result of what. Hopefully, the meaning will be clear.
object economy property unemployment : real object person property unemployed : bool economy.unemployment = COUNT( person P such that P.unemployed )
Now let's write some rules about unemployment:
person.unemployed may cause person.skill decrease person.unemployed may cause person.stress increase ...Skill is desirable and decreases; stress is undesirable and increases; so both of these are costs. Can you give me any other rules of this type?
To make this work we should, in our pseudo-code, note that people have the skill and stress properties, and that one is desired (in general) and one isn't. Then later, we shall be able to automate these rules, by triggering them as the economy changes, flagging changes in various properties, classifying them as desirable or not, and fetching various stories accordingly.
object person property unemployed : bool property skill (desired by self) property stress (not desired by self)
Now, still on unemployment, let's go to loss of output. We can write a macroeconomic rule easily enough:
economy.unemployment increases may cause economy.output decreaseHowever, I want to continue our person-centered approach, viewing as much as possible as an episode in the drama of somebody's life. The economy is not a person and can't suffer, so we can't use a rule like the two unemployment ones above. But is there anyone who very directly deems output desirable? Yes, probably the Chancellor and the head of the CBI. So let's add:
object Chancellor is a person desires economy.output high object Head of CBI is a person desires economy.output highNow our rules will, when triggered, generate episodes in the lives of these two people. These episodes can be linked to stories or bits of stories. Can you provide any other rules and desirable or undesirable properties of this kind to illustrate costs? You may want to add characters to the cast list started earlier. This is an alternative approach to the use of the happiness index: once I've collected enough examples of each, I should be able to make them work together.
That's one way to illustrate loss of output: it helps teach the student what different people in authority deem important (which is related to their ideologies, as mentioned earlier). But we also want to take a more directly person-centered approach, showing how loss of output evolves at the micro level. There are various ways to do this; here is a reasonable one.
object Six Bells is a pubThe idea here is that if someone becomes unemployed, they have less money, so they use their local pub less, so its income decreases. (You could substitute cafe', or music shop, or various other businesses.) Two of the rules are applicable (other things being equal) to most people and pubs; but to focus attention on individual consequences, I've made the second specific to a particular person and pub. This will force the story associated with it to be more specific, and hence more effective.
object John is a person person.unemployed may cause person.income decrease John.income decrease may cause John.use.Six Bells decrease person.use.pub decrease may cause pub.income decrease
Now, the simulation will trigger these rules when someone becomes unemployed, and we can link each one to a suitable story fragment. (I am using the single-dot notation person.income to denote a property of one object; the double-dot notation person.use.pub indicates a relation between two objects. This is as complicated as I need to get.)
Can you add any other objects and rule sequences of this kind to demonstrate macroeconomic processes?
Now let's switch from the costs of unemployment to the costs of growth. Consider increased traffic. What does this mean at a personal level? Various things: one is noise, which may disturb sleep. So we'll introduce an object to denote the traffic passing a person's house
object traffic_past_houseAnd a rule:
economy.growth increased may cause traffic_past_house.noise increasesAnd the addition
does not desire noiseto John's properties. Actually, we could handle this in a more general way by having a "national traffic" object which is an aggregate, and defining traffic_past_house to be part of it, so that if one increases, the other does. However, the simpler method will do for the moment.
We could exemplify pollution and other externalities in the same way. So can you fill in some more objects and rules of this type? You may want to add people to the cast.
Finally, what about inflation? Like growth, inflation can not be posited of an individual person. But we can still illustrate it at the personal level in the way it affects savings:
object savings property real_value economy.inflation increased may cause savings.real_value decrease John has savings object person desires savings.real_valueand the way menu costs can occur:
object Landlord is a person economy.inflation may cause Landlord."Stand on a table every thirty minutes and shout out new beer prices".In this last rule, I have, for the moment, not tried to analyse the ontology of what the text describes, but just included it as a whole. If you can fill in more rules of this type, I can make an attempt at analysis if it seems worthwhile for the simulation when I have more episodes.
[ Jocelyn Ireson-Paine's Home Page | Free software | Publications | The Lives Behind the Numbers On the Screen: Illustrating the Social Consequences of Economic Change By Telling Stories On the Web ]