@2WL: Okay, an example of a question:

How does the practice of buying a company and selling all its property (and similar practices) affect the living standards of employees?

Not perfect, but something in that direction I'd say.

Ah, now that's a completely different question. Not fraudulent at all, but it is a fairly common business practice that usually decimates the standard of living of employees.

Its also not something that banks do. Its something that hedge funds or other companies do.

As a number of people have pointed out, defining your question such that it can be answered is important. I would also offer that it is important in the first place to be very careful not to make the question too broad, because the assumptions you make will render any statistical analysis either impossible or meaningless for example:

"What is the death count of Wall Street?" cannot be reasonably answered without also analyzing every impact of every company that trades there, has borrowed or lent money at any time, or has sold a product to a person or company that has done so. Additionally, such analysis would need to include all impacts, both positive and negative; how many lives were saved by products developed and commercialized, how much has life been improved by the economic activity enabled by Wall Street, what is the metric for measuring the improvements in standard of living that have occurred since the middle of the 1800s? All these things need to be addressed in finding an answer to the question you asked in the first post (which as another poster has mentioned, seems framed in such a way as to already have an answer in mind, and to be based largely in politics, rather than economics. And while economics is a truly tricky field to accurately quantify, it's got nothing on political questions)

I can see two much better ways to go about asking your question. One is a very specific microeconomic question, and should be answered with concrete case studes: "What, historically, has the impact on life expectancy been when companies engage in [Insert Business Practice Here]?" You can make a good case for looking at fraud, embezzlement, hiring overseas sweatshop labor etc, but should pick a specific behavior or two so you can study individual case examples, such as suicide and welfare rate changes among Enron employees, or the difference in workplace accidents/fatalities/suicides between the overall US population and, overseas sweatshops. This has the advantage of being specific, and more easily attributable to the people that actually did something unethical. It has the disadvantage of narrowing the scope of your question down significantly, and a lot of specific data may not be available to you.

The other is to broaden your question into a valid macroeconomic question (or pair of questions, really), such as "What is the impact in lives and standards of living of a recession?" and "In what ways do financial and government institutions contribute to the causation and severity of recessions?". The strength of this question is that there is a ton of data out there relevant to the question you're asking. It's also a topic very worth talking about: there are a lot of things we do to cause this sort of economic hardship, and not all of it is as simple as "Wall Street took money and caused problems", and I think discussing this honestly is a good thing. The downside is that reasonable economists can and do disagree on the answers to these questions.

Uhh fair point, I guess that would be a useful study too but it starts to deviate quite severely from the original intent. Not sure if that's a bad thing, this might be a good start when mapping the negative impact of Walk Street..

The original thought was more concerning things like fiduciary managers betting on money loss of their clients, while they tell their clients that they want the best for them along with all those words that evil companies use these days: integrity, transparency and so on.

I suppose a concrete question is quite tricky there though.

That was to goldfinger.

Good post Scurvy Cur.

And I know I 'framed' the question, but that was a question to the forum. I obviously wouldn't use that in an actual research and I would try not to cherry-pick information etc.

"The original thought was more concerning things like fiduciary managers betting on money loss of their clients,"

To know that, you would have to look at their trade books. To do that, you'd have to be a bank regulator. That kinda stuff isn't exactly disclosed in annual reports.

Now you're asking questions about information that is going to be difficult or impossible to obtain, much less measure. Statistics on corporate fraud isn't exactly something that accounting departments track in a spreadsheet. Like Scurvy Cur said (in an excellent post) economics is a tricky field to quantify.

In economics, a lot of studies deal with either natural experiments or proxy variables. You might not be able to find clear statistics on corporate fraud, but you could find a statistic that approximates corporate fraud. For example, you could examine a relationship between corporate fines levied by the SEC and bankruptcies declared in the US. Corporate fines *could* be a proxy for fraud, and bankruptcies *could* be a proxy for "negative impact of Wall Street", whatever that means.

Steephie, I agree with you that Wall Street and its practices have killed many, but you are clearly leading into this with some rather poorly veiled bias. There's no way that anything you find is worth much if you already know the direction that you are headed.

No one will ever admit the evils and deaths caused by capitalism - just like if communism had won the cold war, we wouldn't be able to talk about the millions who died because of it.

But you are talking about both right now....so....irony?

@bo: I believe I'm capable of doing the actual study objectively (as far as possible). I'm not exactly nuancing my thoughts here, but I'm genuinely interested in how big a deal it is. Are deaths and lives ruined incidental or structural? Are there a few hundred lives ruined per year in the whole industry or can they compete with ISIS in structurally killing people, though hopefully not on purpose?

I don't think anyone knows the scale and severity. That is something I would like to investigate in an objective manner. Then, people can start thinking about what and how much should be done, if anything.

If I were to look at the annual profit of fiduciary investors and the annual profit or loss of clients, would that be scientific? If I find, for instance, that there's a relation between how much the investor wins and how much we lose, could I conclude from that that, according to economic principles, the investor will want to optimize his profit which means 'optimizing' the loss of their clients?
Both the variables I'm using should be in annual reports, right?
From there on, could I estimate how much damage to people/the economy is done per dollar earned this way, and in conclusion combine those findings and give an estimate of how much damage they cause on the whole?

I'm not sure whether that's scientific, though..

....but how would that lead you, specifically, to deaths?

Association is one thing causation is another. You need a very limited number of variables, lots of samples and a null hypothesis which you will either support or not. Note you will never be able to prove what you suspect in absolute terms.

@Jamie: Not. I'm considering options.

@fulhamish: Which step are you referring to exactly?
1. Looking for a relation between the loss of clients and the profit of fiduciary investors.
2. Concluding that the most profitable thing for the investor to do typically causes a loss for clients.
3. Estimating how much damage is done to people/the economy per dollar
4. Combining the estimated damage per dollar and the amount of dollars involved in the companies researched to come to an estimate of how much damage is done
(Possible) 5. Extrapolate the results to more companies, since I would probably only scrutinize a sample size.

Steephie, it is really good that you are interested in doing this work. I think, however, that there is a big danger that you are being over-ambitious, that is if you want robust quantitative rather than qualitative results. For example, it is well known that Merrill-Lynch etc. busily promoted investments to a set of retail investors, while at the same time shorting the investment themselves and on behalf of a more select group of clients. I know it is bull-shit, but they justified this by saying either that they were ''making a market'' or, slightly better, that they had erected robust Chinese walls within their organisation.
In effect your points 1 and 2 have therefore already been established. Your points 3 and 4 (5) are really much too nebulous for a quantitative study. You need to pin things down to a clear null hypothesis (I gave an example earlier, you might think of an alternative) and then set out to support it or not by statistical analysis. Note if you suspect that these institutions do indeed act in a detrimental way to the wider population, then I think that taking as large a sample number as possible would help. Beware however of the temptation of data mining – your results should be repeatable to another researcher. In other words be fair and start with a clear hypothesis and set out to analyse it, one way or the other.

On the other hand, having said all of that, you might be happy with a qualitative approach; if so all well and good.
I hope that helps, but I am not a social scientist or economist, somebody else might help to tune your study to those disciplines.

What would a qualitative approach be like? I'm not native English, maybe that's part of the reason I'm slowly starting to not see what some of the things you say mean in practice.

I think you're also working within the realms of social constructs. You're looking at the behaviour (corporate) and then specifically naming those behaviours. It's possible to run with some quantitative work but you're so dependent on definitions, etc.

Do you mean I am specifically being too dependent on definitions etc. or are you simply stating that's an issue as well?

Verbal description, rather than statistical analysis or 'proof'.

Right, so I would just investigate the details of how it's done and perhaps make clear I'm not talking bullshit by pointing out an example?

I think that would have little.. weight. Besides, how is rather straightforward and not surprising for anyone who knows anything about it.

Simpler but less robust might be to just correlate stock market moves with death statistics country by country. It might not be easy to obtain a lot of the data and you might be end up finding a correlation between an increase in ice cream sales and people drowning.

It's a known fact that eating icecream causes death by drowning.

All this talk of over ambitious ness makes me want to say: if you are not necessarily trying to publish results in a rigorous academic setting but just want to learn about the works for yourself and maybe share the results online, I say go for it and try to perform a rigorous analysis.

For example, in college I wanted to know: what is the correlation, quantitatively, between war and hunger? I used a dataset listing conflicts and wars in all countries year by year with an intensity rating based on deaths and damages and displacement and took the upper end of the scale as bona fide warThen I looked at countries with a malnourished population exceeding some benchmark, I think 10% or 20% of the total population. Among these war torn and hungry nations from 1990-2010 I created a study group and then found a control group of cohntries with the same levels of hunger but without the war in the same period, and controlled for the factors of GDP per capita, ecological setting, political region (west africa, South Asia, etc) and a couple more I can't remember, like corruption and population density or sone thing. Then I peformed a statistical analysis and found that the présence of a war 1990-2010 correlated with a 20% increase in the number of hungry people.

This isn't some scientific law, but it was very educational for me. I now have some inkling in quantitative terms of what the consequences of an outbreak of war could be, that, or what the consequences of pervasive hunger are. It's significant. 20% means millions of people. It was one more to thing that convinced me ever more that global hunger should be the top priority of my life.

Learn about the *world for yourself I mean

Well, the thing with that is that correlation is not causation. Maybe war causes hunger, but lack of hunger does nothing to prevent war. Maybe a third factor causes both while they don't influence eachother directly.

I get your point though and it is a useful study, if only so you know about/can point out the correlation. Then you can either draw reasonable but unscientific conclusions or proceed to look for ways to figure out what causes what.
Maybe follow the countries for an extended time-period and see what happens first: the outbreak of war or an increase in hunger. Then you may find a pattern..

Back to my specific case though, I think the real problem is finding the data. Wall Street knows how to cover up shams pretty well. Perhaps I could look into the number of instances uncovered in different periods, but I don't know what I could conclude from data found there. Did they get more reckless/careful with their cover-ups? Do people try to keep them in check more or less? Did the number of instances simple increase/decrease while the amount uncovered remained the same?

Either way, I would only find a trend rather than a sense of scope, I guess.

It's effectively the same thing as with ordinary crimes: finding out how many people were prosecuted should be pretty easy, finding out how many crimes are committed is pretty much impossible.

I do admire the intellectual curiosity! I am not sure how many of us keep that up post-formal ed, unless actually employed in research.

I'm actually in high school (though I'm reaching beyond high school in my way, I think it'd be unfair to say I'm at 'high school level' and I think a fair amount of teachers would agree with me). It's not even a particular interest for research, I'm just always looking for new projects, whether that be writing a book (work in progress), creating a site (check, twice), creating a boardgame (work in progress), starting a bussiness (work in progress) or this.
Perhaps I can make a living on such projects. Most would say that's ridiculous but you could safely say I'm betting on multiple horses. If I bet on a new horse once or twice a year, I've been betting on more than a dozen horses by the time I'm done with a master.

So yeah, that's a metaphor for over a dozen projects finished in my early twenties, hoping that there's enough money there to keep me from any regular job. If not, oh well. I can live fairly minimalistic though, and that's definitely a sacrifice I would make for not having to live a 'normal' life :P
I just enjoy projects. A purpose, something to get me in the zone and something that interests me all in one! I like being useful and I estimate my will grow towards helping people more and more as I get older, it's just tricky to make a difference in that regard when you're in high school I guess, so I'm keeping it slightly more realistic.

What? You were only admiring my intellectual curiosity?

I never said I showed anything other than correlation, my point was that although I had a limited and flawed result, I knew more when I was finished than when I started and I learned something that wasnt published elsewhere - I found it out for myself. So yeah I think you should do something similar. If nothing else you get an idea of the order of magnitude you are dealing with in terms of casualties of capitalism. It's a great topic. I often ask myself how many people die as a direct result of my country's inadequate healthcare system. I couldn't even guess