Wingnuts "R" Them...

Ummm... Ezra:

Ezra Klein: Kashing In: And speaking of the overrepresentation of libertarian econ bloggers, I was interested to see this study (pdf) showing that precious few economists are actually free marketeers. Indeed, only a couple of percent qualify...

Here's the "study":

http://www.sofi.su.se/wp/WP06-6.pdf Is There a Free-Market Economist in the House? The Policy Views of American Economic Association Members. By Daniel B. Klein and Charlotta Stern: People often suppose or imply that free-market economists constitute a significant portion of all economists. We surveyed American Economic Association members and asked their views on 18 specific forms of government activism. We find that about 8 percent of AEA members can be considered supporters of free-market principles, and that less than 3 percent may be called strong supporters...

Here are a few choice words from one of the authors of that "study":

http://www.scu.edu/civilsocietyinstitute/events/upload/DiscriminationKlein.pdf: Antidiscrimination Laws Violate Freedom Maryland used to prohibit whites and blacks from playing tennis together. H.L. Mencken protested:

Is such a prohibition supported by anything to be found in common sense and common decency? My answer is a loud and unequivocal No. A free citizen in a free state, it seems to me, has an inalienable right to play with whomsoever he will, so long as he does not disturb the general peace. If any other citizen, offended by the spectacle, makes a pother, then that other citizen, and not the man exercising his inalienable right, should be put down by the police.

But if freedom means being allowed to play tennis together, it also means being allowed to choose not to play tennis together....

Along with self-ownership, the freedom of contract (or of association) is a cornerstone of freedom or liberty, as understood by the whole classical-liberal tradition....

1. Vague, complicated govt rules are imposed from the top-down. They lead to uncertainty; people don’t know what’s unlawful. People are confused about appropriate behavior; social norms of esteem and reputation don’t work well when people don’t know what kind of behavior is “bad.”
2. The enforcement of interventionist rules falls on government.
3. Specifying an Anti-Discrimination Law. In formulating an anti-discrimination law, a philosopher-statesperson must specify:
   • What categories are protected
   • What constitutes belonging to the protected category
   • In what activities such discrimination will be unlawful
   • How the law will be administered and enforced
   • What kinds of behavior constitute unlawful discrimination....

Civil Rights Act of 1964 reads: Section 703.

It shall be an unlawful employment practice for any employer to fail or refuse to hire or to discharge any individual, or otherwise to discriminate against any individual with respect to his compensation, terms, conditions, or privileges of employment, because of such individual’s race, color, religion, sex or national origin....

Intervention presumes knowability (Hayek). But the relevant social facts are usually not knowable. Who knows what lurks in the hearts of men? Only the Shadow knows!... People become pitted against each other, suspicious....

Sometimes antidiscrimination laws harm their purported beneficiaries.... One never knows whether someone attained a position because he was a member of a protected category. The only way you can know that you can ride without training wheels is to ride without training wheels.... Antidiscrimination laws demean members of the protected categories.... By coercing ordinary citizens, antidiscrimination laws might make the disabled swell with a sense of power, but the law also sends the message that without government privileges they cannot assert their own worth and dignity....

Antidiscrimination laws create backlash. People resent being forced, they develop hatreds. One reason racial and gender issues are so bitter and acrimonious is because force is involved....

The last of these is, I think, the ugliest.

Only Daniel Klein thinks that the class of free-market economists is identical to that of ignorant, vicious wingnuts.

Comments

[Ezra]

That's certainly ugly. But I read the study and didn't notice any similar sentiments infecting the methodology. As I said, I think it's a bit oversimplified, but I don't think the study was racist.

[howard]

juxtaposition: do you know anything about the nature of hiring discrimination? do you know anything about the way things were in america from 1876 - 1964? do you know anything about the utter waste of talent and capability because of people like you who confuse freedom of association with discriminatory hiring practices? do you know anything about the number of incompetent white males who were promoted over the generations, accruing position and capital for their families and thereby providing their offspring opportunities denied to those who were discriminated against in order that incompetent white males could be promoted?

[Pithlord]

That's just standard-issue libertarianism on anti-discrimination laws. All your libertarian friends think that way.

[My libertarian friends are smarter than that--they have some knowledge of social conventions and social pressures.]

[David]

It's hard to believe that they would believe a "free-market advocate" would put '5' for all of those issues.

[howard]

C.L. Ball, if the market on its own were going to achieve non-discriminatory nirvana, it would have done so well before 1964.

maybe in a parallel universe, where there is a real free market and oligopolistic power doesn't develop and create political support for entrenched power, we could count on a market solution to take care of discrimination, but in this universe, we couldn't.

after all, if everything in the entire universe reinforces to you, as a business owner, that (for example) blacks are inferior to whites, how could you possibly imagine that non-discriminatory hiring practices would improve your business position? And even if you did, on your own, in the face of social reinforcement to the contrary, develop appropriate hiring practices, the other discriminatory firms would preclude you from reaping the advantages of superior talent (after all, how could you responsibly hire a black sales person if you knew that many households wouldn't answer the doorbell? how could you responsibly hire a black senior executive if few hotels and restaurants would allow him or her to stay or eat there?)

now, as for the ADA, i agree, the law is very poorly written and has led to some important social improvements and some real silliness and unhappiness. that, of course, is not a problem with the 1964 civil rights act.

[steve]

In a capitalist economy, the promise is the possibility of wealth and certainly a living wage if one works hard, takes risks, etc. To undercut the economic promise of a capitalist system because of race, creed, color, gender, etc is to undercut the promise of the capitalist system that has made it so effective in growing economic wealth.

On the point that anti-discriminatory hiring practices endgender anger and the feeling that the recipients are less worthy. The anger comes because the club that allowed one group to maintain control over hiring for themselves, like the Irish in the NYPD, is no longer. They can't discriminate anymore, too bad. They built their castle out of sand.

And as for set asides, etc. in college admissions. If you are against quotas, etc., but think there is no problem with George Bush going to Yale even though his grades and test scores would never have gotten him, then you are fool.

Want to live in a caste society perpetuated by law? Move to Calcutta -- as an untouchable. Let me know how it works out.

[Colin Danby]

I'm reminded of zealous vegans who insist on appropriating the label "vegetarian" for themselves on the grounds that cheese-eating vegetarians aren't True Vegetarians.

One awaits a reference to support "Adam Smith’s way of seeing the government as essentially
coercive and as riddled with pathologies" (25).

Anyway the point, Ezra, is not that the study is inherently racist but that the authors set it up so that you have to be a pretty hardline libertarian to qualify as "free-market." LOOK at the questions they use to make the index.

[dale]

The phrase "self-ownership" is an ugly metaphor. I suppose it means autonomy but the author felt the need to subsume a human quality to some sort of market status.

[rea]

The "No True Scotsman" fallacy in action!

Argument: "No Scotsman puts sugar on his porridge."
Reply: "But my uncle Angus likes sugar with his porridge."
Rebuttal: "Ah yes, but no TRUE Scotsman puts sugar on his porridge."

http://en.wikipedia.org/wiki/No_true_Scotsman

[Mike Huben]

In our way of thinking, being a free-market supporter is about favoring the principles for social rules that imply a free market. It is not an issue of the principles of economics. A “natural rights” libertarian who has no economic understanding is nonetheless a supporter of free-market principles.

In other words, this is simply a libertarian purity test.

However, the objective (unlike the Nolan quiz) is to chastize for not being pure enough. Hence, the scores are judged to be low, instead of inflated as in the usual quiz.

[Mike Huben]

(Gack: my italization of the first paragraph was stripped.)

The first paragraph of my previous message was a quote from the study.

[pgl]

There were really two parts to Ezra's post - the 1st being something I highly recommend. That being putting Kash's new blog in one's blogroll.

[AAustin]

Let's try a different tack here. Natural rights and market efficiency do not always play well together.

Markets may fail to be efficient because of externalities, among other things. All competent economists acknowledge this, tho may disagree over the importance of specific externalities. For example, a mixed-race tennis match may bother a third party, which constitutes an externality. Whether one regards that externality as serious or not is a judgement call.

Amartya Sen constructed a simple example which demonstrates the impossibility of reconciling (Pareto) efficiency and liberalism, meaning that one has autonomy to make at least some decisions which affect oneself without outside interference.

Lewd would rather read a dirty book rather than not read it. But he would even more like to have his colleague Prude read it. Prude would rather have no one read the dirty book, but in the spirit of sacrafice, would rather read it himself rather than let Lewd read it. Having Prude read the book is efficient, in than moving to any other outcome makes someone worse off. However, having Lewd read the book is the only outcome consistent with a minimal degree of liberalism.

The point is that there are good reasons to ignore certain externalities that interfere with others' ability to make autonomous choices. In that sense, Mencken in the quote above has that right: one should be free to play tennis with whom one chooses, regardless of who that bothers. On the other hand, Milton Friedman is wrong when in Capitalism and Freedom argues that store owners should be able to discriminate in hiring, not because of their own prejudices, but because of the prejudices of their customers. There are good social reasons to have owners ignore externalities which result from the prejudices of customers.

[PaulC]

"Sometimes antidiscrimination laws harm their purported beneficiaries"

It always amazes me that people (particularly economists) get away with applying ridiculous standards to things they oppose.

It's true that "sometimes" a law may have unintended, detrimental consequences. The big question is what does it do most of the time, and how does that balance out not having the law. By applying that (reasonable) standard, it's clear that the Civil Rights Act has been a resounding success, and it's hard to find even a conservative claiming otherwise outside of wingnut circles and thinktanks.

Did you know that sometimes having a car actually hinders rather than helps your ability to get from place to place?

Did you know that sometimes taking a prescription drug under advice of a competent doctor actually makes you worse off than you were to start with?

Did you know that sometimes inhaling the surrounding air will kill you?

Taking "sometimes" to mean "there exists a set of circumstances that I can present as a hypothetical" you can claim that anything sometimes has the effect that you would like to insinuate is its primary effect. I think that these kinds of "sometimes" statements are a useful red flag in distinguishing polemic from reasoned analysis.

[Ken Houghton]

My economics teacher in B-School used the example of 9 "A" guys and one "B" guy all applying for a job. In the world in which you must not discriminate against B, hisser being tied with one of the As as most qualified will lead to hisser being hired.

So 9 "A" guys will now resent that they were passed over for the "B" guy.

This apparently is a much different situation than if there had been 10 "A"s applying for the job and nine of them lost out.

[Colin Danby]

This is quite reasonable AAustin and what jumped out for me in the survey is the usual homily about externalities. It shows a group of people quite strongly opposed to tariffs and state ownership of capital, but prepared to consider public ed, environmental regulations, and so forth.

To be fair to libertarians, I think for them efficiency is not the key criterion -- they see gov't interference as deeply wrong in itself and are happy to put up with inefficiency to keep the state off their backs. And a variety of objections can be made to the whole static efficiency approach.

Ezra's categories of "liberal" and "conservative" in his original post are possibly not coherent.

[PaulC]

It also occurred to me that I don't know why you'd think an economist is a "free marketeer" anymore than a meterologist would be an advocate for refusing to carry an umbrella or purchase flood insurance.

Any competent economist can be expected to understand the power of markets, and will probably not suggest that better efficiencies would be achieved through a centrally planned command economy. But it doesn't mean that all or even most economists would claim that the market optimizes everything in exactly the way that everyone wants.

I agree with the comment above that libertarians are not looking for some kind of theoretical optimum except by claiming as a matter of definition that one must minimize state control. Most people, including most economists, have a more pragmatic view of constraints and objectives, and only a very naive person would propose that a completely unregulated market inevitably leads to all the best trade-offs.

As a society, we make conscious trade-offs, and as long as we reject the libertarian notion that regulation is bad by definition, the question is how to balance regulation and the distributed decision-making of the market to realize our chosen trade-offs as closely as possible.

[C. L. Ball]

"C.L. Ball, if the market on its own were going to achieve non-discriminatory nirvana, it would have done so well before 1964."

Klein's slides do not argue that the market alone would have created non-discriminatory nirvana (and neither have the civil rights laws due both to under-prosecution of violations of the laws and to what those laws don't cover, such as political discrimination).

Klein argues that if state-mandated discrimination was barred _and_ society-imposed discrimination was permitted, market dynamics would crowd out irrational discrimination. This would be an alternate post-1964 history in which only the state-mandated discrimination ended.

If social discrimination is more powerful than market dynamics, then the CRA should have had only marginal effects in many areas. Take the black salesman example above: since the CRA, a firm hires the black door-to-door salesman if he is best qualified for the job. But if bigoted customers still refuse to buy from the black salesperson -- the CRA does not compel consumers not to discrimnate -- then the black salesman won't sell much. He'll get fired for low sales because he is black but not because his firm discriminated against him because he was black. Why didn't this happen in fact? Because social discrimination was not that powerful; people would buy from black salesmen. But pre-CRA, there were many laws that banned non-discrimination. If social discrimination were overwhelmingly powerful, bigots wouldn't have needed the laws mandating discrimination.

Klein's argument breaks down for semi-rational discrimination (e.g., "I don't want to hire a person in a wheelchair because I might have to build a ramp, if it is affordable, to accomodate the person, and I don't want to spend the money on that"; never mind that any existing employee who is temporarily wheel-chair bound would benefit, or that ramps make it easier to move heavy objects).

The ADA is a good law for the most part since it addresses semi-rational discrimination that can cause market failures. The flaw is that costs of accommodation are only partially socialized (specific employers pay) rather than being fully socialized (governmen pays). This in turn has led many institutions violate the spirit of the law (e.g., universities that shift the compliance burden onto individual employees rather than developing adequate institutional responses).

[Bernard Yomtov]

"Klein argues that if state-mandated discrimination was barred _and_ society-imposed discrimination was permitted, market dynamics would crowd out irrational discrimination. ."

Klein is badly wrong. There were many areas of American life where discrimination was not mandated by law, but was prevalent. And does he accept the "rational" discrimination of the type PaulC says Milton Friedman approves of? I do not.

In fact, the general utterly-out-of-touch-with-reality thrust of libertarian arguments on anti-discrimination laws is one thing that makes it impossible for me to take libertarianism seriously. My impression is that libertarians think of racial discrimination as a sort of bizarre idiosyncracy that we should shrug off in the interest of indvidual liberty, rather than a systemic pattern of mistreatment of a sizeable segment of the population.

That anyone can oppose efforts to eliminate this mistreatment on the grounds that forcing an employer to ignore his prejudices and hire some black people, or forcing a restaurant owner to serve blacks, is too horrible to contemplate, is ludicrous.

[andres]

The question of how narrow the definition of "Libertarian" should be is answered in part by one's overall sympathies to the contemporary Libertarian agenda.

Those analysts hostile to Libertarianism tend to adopt a very narrow, "Type I" definition, eg a Libertarian is someone who dogmatically favors liberty as an absolute principle regardless of ethical, efficiency, or utilitarian concerns. Under this definition, Libertarians would include only those fanatics who oppose anti-discrimination laws and even market-based pollution regulation measures such as tradeable emission permits. Not a very flattering picture--I wouldn't be surprised if Milton Friedman today secretly regards his chapter in _Capitalism and Freedom_ on discrimination vs. the market as something of an embarrassment.

Some definition of this sort is what was used by the Klein and Stern paper.

On the other hand, broad sympathy with Libertarianism, though not agreement with every single one of its positions, would tend to define Libertarians under a "Type II" rubric as people who generally favor free market, non-regulatory solutions unless a good logical and empirical case can be made to the contrary. Most economists, including Brad, are by this definition Libertarians in spite of there being quite a few cranks who give Libertarianism a bad name.

The Type I definition has won out in the minds of most intelligent observers as well as (surprisingly) of Libertarians themselves, which is imo a good thing since it enables me to immediately dismiss any self-identified Libertarian as a cultist crank. Not the most tolerant of reactions, but usually one that saves time and patience.

Still, one must be careful of the boundaries used in most definitions. Most people tend to use a similar Type I definition when they read or hear the words "Marxism" or "Marxist", whereas many self-identified Marxists use a broader, Type II definition to define their own outlook, and yet fail to understand this definition gap when, to their never-ending consternation, non-Marxists subsequently react in a hostile manner.

[James Killus]

I find it more interesting that it's always affirmative action where these arguments are used. I wonder what would happen if someone tried to apply them in some other venue:

"Laws that require the washing of hand's by restaurant personnel create a backlash because force is involved. It would be much better to allow market forces to work, and any restaurant that had a hepatitus incident would quickly mend its ways."

"Laws against rape invite backlash and anger against women. There would be fewer rapes if it weren't illegal."

"I'd like to beat the crap out of Daniel Klein. I don't bear him any real animosity; I just don't like being told what I can or cannot do."

Yeah, that works.

[guest]

"Laws that require the washing of hand's by restaurant personnel create a backlash because force is involved. It would be much better to allow market forces to work, and any restaurant that had a hepatitus incident would quickly mend its ways."

No need for a specific law - common-law negligence standards are enough.

"Laws against rape invite backlash and anger against women. There would be fewer rapes if it weren't illegal."

"I'd like to beat the crap out of Daniel Klein. I don't bear him any real animosity; I just don't like being told what I can or cannot do."

Huh? Rape and battery are egregious violations of natural rights.

[Bernard Yomtov]

"in the Becker-type models, we don't really get discriminatory equilibria absent consumers being willing to pay a bit more to avoid being served by someone against whom they harbour prejudice. "

Having lived in the Jim Crow south I can assure you that people were happy to pay more to maintain their prejudices.

While I appreciate your comment, I have to say that talking about what economic models may or may not show in this area strikes me as the worst kind of ivory-tower intellectualizing. If the empirical literature doesn't show what anyone with eyes to see would have known then the methods used have serious problems.

[Eric Crampton]

Bernard: the nice thing about stylised economic models is that we can then compare them to the real world and enquire about the differences between the two. If the model predicts no discrimination absent strong preferences by consumers and we see lots of discrimination where consumers seem only to have weak preferences, we know we need to look at other reasons. One of the other reasons, in the Jim Crow era, was legislation mandating discrimination. Part of the point of the Jim Crow laws was that people wanted to indulge their discriminatory and hateful preferences without having to pay for them directly: that's why we saw legislation requiring discrimination. I'm sure there were employers whose taste for discrimination was sufficient that they'd have engaged in it without being required to (even though it would have been detrimental to the bottom line); I'm also sure it's the case there are some racists who'd be willing to pay a lot to avoid being seated next to or being served by someone. But, in a competitive market, those folks are forced to pay a real price for indulging their preferences. Again, look to the streetcar case. There were tons of racists, but the streetcar company simply couldn't make a profit by running segregated cars: racists weren't willing to pay a sufficient premium for it to be worth the company's while to segregate the cars. Racist voters forced everyone to consume their preferred bundle by mandating discrimination at the ballot box.

Glen Loury's work on statistical discrimination is more relevant to today's policy world. He looks at what happens when, on average, blacks have lower human capital than whites and when employers have rational expectations of the same. In some cases, affirmative action may improve outcomes, but in others, it simply reduces the expected return to human capital acquisition and makes matters worse. Information asymmetry about job applicant quality would seem a more likely mechanism for generating discriminatory equilibria than employer or customer prejudice; the appropriate policy recommendations that flow from such findings, though, are rather ambiguous. One place to start might be in fixing inner-city schools such that employer expectations of human capital start rising. The Loury model provides ambiguous results on things like affirmative action; alternative mechanisms that improve black human capital acquisition, either by lowering its cost or by augmenting its return, would be less ambiguous in result. The "ivory tower" models can be rather useful from time to time.

[C. L. Ball]

" My impression is that libertarians think of racial discrimination..."

Klein's argument was not about racial discrimination but *all* anti-discrimination laws, and in part because the basis for discrimination is often politically determined.

There is no federal protection against hiring discrimination for over-weight people, for example, unless the person is over-weight because of a disease or is perceived to be obese because of a disease. An employer can legally say, "I don't want to hire you because you're fat and I think fat people are ugly and will turn off my customers." A restaurant owner can refuse to seat you because you're wearing a jacket with a picture of Che Guevara or Rush Limbaugh on it.

But if a restuarant owner did not want to seat Southern Baptists because their convention oppose equal rights for gays and lesbians, he cannot because the CRA outlaws discrimination on religious grounds.

[Bernard Yomtov]

Eric,

Thanks for your comment.

A few points I would like to make.

First, you write,

"If the model predicts no discrimination absent strong preferences by consumers and we see lots of discrimination where consumers seem only to have weak preferences, we know we need to look at other reasons."

We don't know this at all. The model may simply be invalid, or our judgment of preferences may be inaccurate.

In fact, I believe there were very strong preferences for discrimination. I don't have a model to prove this, but a bit of historical research might help. Political campaigns in that era, for example, were often contests to see which candidate could be more strongly racist. That doesn't fit neatly into regressions, but it is no less true for that.

The segregation laws, as you point out, did not write themselves. They reflected the will of the voters, by which I mean the whites of course. Nor were they supported only by a relatively small group of racists who wanted to dodge the costs of discrimination. The laws were immensely popular and any southern politician who proposed their repeal would find himself out of office, or worse, very quickly. So I don't agree that

"Part of the point of the Jim Crow laws was that people wanted to indulge their discriminatory and hateful preferences without having to pay for them directly: that's why we saw legislation requiring discrimination."

Many, a great many, people supported these laws because of plain racism and dislike of seeing races mingle. To look solely to economic motivations is a mistake.

In addition, I think the notion that discrimination is necessarily economically costly is wrong. It is much too facile an assumption. In general, discrimination was economically beneficial. (I'm not familiar with the streetcar case, so that may be a counterexample) Assuming the absence of Jim Crow laws, a southern restaurateur who served blacks would go out of business quickly. An employer who hired blacks for responsible positions was going to have lots of trouble with his work force.

To argue that this employment discrimination was harmful to the employer is to confuse, as many who make this case do, productivity with characteristics like intelligence, diligence, etc. The most conscientious worker in the world will not be productive if he has to deal with contempt and lack of cooperation from coworkers.

My point here is that the social context of the segregated south was such that many of the implicit assumptions about economic behavior - such as "adding a customer is desirable" - do not hold. One of my complaints is with those who use refuse to grasp this, and take the model as more accurate than reality.

You write,

"in a competitive market, those folks are forced to pay a real price for indulging their preferences. "

In the context I describe this is not true. Consider Major League Baseball as a "laboratory experiment". Teams were all white until 1947. Why? They were not, in general, under legal constraints. They could have signed black players. Further, because of the Negro Leagues, the identities and skills of many excellent black players were known. Either the owners were fierce bigots who let racism overcome their wish to improve their teams, or they felt that it would be economically unwise to sign black players. In fact, some owners expressed the latter sentiment. In any case, the notion that competition wipes away discrimination, or does so in a reasonable time frame, does not seem to hold in this instance.

[Bernard Yomtov]

Cl Ball,

"If social discrimination were overwhelmingly powerful, bigots wouldn't have needed the laws mandating discrimination."

But that doesn't mean they wouldn't have passed them.

[James Killus]

Not all people are as irony and logic impaired as "guest" (not his real name), but I think I'll make my point more clearly.

Consider this argument:

"Tax cuts for the rich can cause a backlash and resentment against the rich and that will harm their interests in the long run."

Some might agree with this argument; some might disagree. However, I think it's pretty clear that it can serve as something of a litmus test. Few, if any, are going to agree with it if they are favor of upper income tax cuts. In other words, the likelihood of believing a marginal argument depends strongly upon whether or not someone agrees with an underlying policy agenda.

In the case of affirmative action, many of the stronger arguments against it are also arguments against much more important examples of discrimination, like those of birth, wealth, social class, etc., yet discrimination that is perceived as being "against whites" yields far more emotional vehemence, while discrimination against African Americans and other darker skinned minorities tends to be shrugged off. Those who are against affirmative action first need to address this particular fact.

While it's certainly not true that those who deride affirmative action are white supremacists, I can't imagine that there is a single white supremacist who is in favor of it. While "politics makes strange bedfellows," it's worth noting that if you find yourself in a strange bed, you might want to think about how you got there, and why your bedfellow finds you attractive.

[James Killus]

That last should read "While it's certainly true that not _all_ who deride affirmative action are white supremacists..."

[guest]

I can't imagine that there is a single white supremacist who is in favor of it.

I would imagine that most "white supremacists" would be in favor of so-called "affirmative action" in their *own* favor. (Public Choice 101: concentrated interests trump diffuse concerns). You may want to reconsider your argument.

[guest]

"discrimination that is perceived as being "against whites" yields far more emotional vehemence,"

Whether affirmative action policies are "for" or "against" whites is an irrelevant contingency. The point is that government-mandated discrimination is open to serious abuse. "Market" discrimination may not be self-correcting, but at least it's self-limiting.

[James Killus]

"Guest,"

Neither your logical skills nor your knowledge base seem to be improving. Give it up.

[Walt]

The weirdest thing is when people quote economic ideas like they're some sort of gospel. Economics provides powerful tools to understand the world, but it doesn't do it all by itself.