You are viewing marknau

marknau Below are the 10 most recent journal entries recorded in the "marknau" journal:

[<< Previous 10 entries]

April 26th, 2009
10:31 am

[Link]

Parenting with ECON
The younger daughter isn't feeling well today, so the daughters are going to stay indoors and watch a movie. How do we determine which movie is going to be watched?

One daughter selects X movies from our collection. The other daughter picks one of the X to be watched. We conduct an auction to determine which role is filled by which daughter. Both wanted to be the initial selector all the way up to 8 movies, at which point the younger one decided she wanted to be the one picking the final movie from the array of 8. So that's how we did it.

"The Incredibles," in case you were curious.

(4 comments | Leave a comment)

March 12th, 2009
01:45 pm

[Link]

Macroeconomics: The Numbers Racket
How well off are we, in aggregate? It’s an impossibly complicated, possibly unanswerable, question.

Let’s say you had omniscient power, and could instantly and unerringly comprehend all of the things that people were doing, how they felt about their lives and their experiences, the material possessions they have accumulated, the relationships they’ve formed. Even then, how do you score all that?

And whose scoring method do we use? I personally think that every Brussels sprout that has ever been grown makes each of us a lesser person. That guy over there thinks that burning corn husks on an altar to Loki is the only valid form of human expression. How are we to place a value on all of the things humans are doing with their lives, and to sum it all up?

You can’t. And you can’t even get a good approximation. What you can do is get a really arbitrary, error-laden, wholly unreliable number that might give you some limited insight, in a very hazy way, into this chaotic panolpy. You do this by basically adding up all the money that people pay for the goods and services they receive from other people. Let’s call this our HWE (hand-wavy estimate.)

Wait. What? A person who loves to cook spends a great evening making a delicious meal for his family and neighbors. Score: Zero. Doesn’t count. Well, we’ll count the price of the ingredients . But nothing for the labor. And no score for the time spent together. So far as the HWE is concerned, you might as well have served the ingredients raw and eaten them in separate rooms. Madness!

A weekend afternoon spent running around with the kids at the park? Zero. Doesn’t count. HWE takes no notice. Stay-at-home parent cleans the house, organizes the craft materials, and does laundry. Zero. Hire a maid to do all that? Ok, now we’re talking. We’ll add whatever you paid her to the HWE.

Sat in traffic for 2 hours on the way to work? Sucks, right? No subtraction from the HWE for that. In fact, it will probably increase the HWE because you’ll need to use more gas.
Have a job you love? Hate? HWE doesn’t care. Get to spend a lot of leisure time to yourself? Don’t care. Have a lot of friends? Strong community? Don’t care. Doing things for yourself? Doesn’t count. Enriching yourself and others in any way that doesn’t involve a money transaction? Might as well not have happened.

OK, OK. You get it. The HWE is a really poor estimate. “But,” you say, “it’s the only gauge we have. And sure, there is a lot of noise, but overall it averages out, right?”

There’s two major problems with that:

1) People latch onto the HWE as if it were a precise measurement. It is used to make comparisons across time, across culture, and in numerous other places where the noise doesn’t “average out.” The very fact that you put a number on the HWE tricks human brains, even those who should know better, into thinking that it’s a solid model of reality.

2) Short-term swings in the HWE are the main input we use to determine economic public policy and who governs. Which is not only committing a grievous offense of the problem #1 variety, but also encourages the governing apparatus to FOCUS ON MANIPULATING THE HWE.

If modern medicine subscribed to the theory that filling a patient’s mouth with icewater was the proper cure for a fever (it brings the oral thermometer reading down, you see,) we would rightfully decry it as quack science. Yet modern macroeconomic prescriptions can be floated which are perfectly analogous to the icewater treatment, and they are treated as being entirely respectable. “Bring that HWE up! Whatever the cost!!!”

Speaking hopelessly simplistically for a moment, the way we spend our time, intelligence, creativity, energy, and resources can be put into three categories:

GROW: This activity isn’t particularly valuable right now, but will pay off over time, helping future people be more efficient or effective many, many, years down the road. You’re researching a cure for cancer, manufacturing drill presses, teaching mathematics, building a highway.

USE: This activity is aimed toward near-term enjoyment or usefulness. You’re enjoying a day off, making television sets, policing a neighborhood, visiting family, flipping burgers.

WASTE: This activity, on net, brings no joy, utility, nor anything positive into being. You’re doing a job you hate making things nobody wants, miserably unemployed and not enjoying your leisure time, or spending your considerable mental talents figuring out how to swindle people out of their possessions.

In the messy world of non-theoretical reality, there are no pure examples of any of these categories. For example, A very smart and well-educated lady who is spending her time lobbying Congress is engaging in WASTE to the extent that she is simply counter-acting the efforts of some other lobbyist, engaged in some GROW and USE to the extent that she is helping the decision-making process work more accurately in deciding common resource allocation, some combination of USE and WASTE when she attends networking events, etc. Just about every activity will be some mix of all three.

Again roughly speaking, the more we have toward USE, the better we are currently living. The more we have toward GROW, the better people will live in the future. WASTE has no upside, but will always be with us so long as we are imperfect beings.

Now, a grave problem with HWE is that it can’t distinguish at all between GROW, USE, and WASTE activities. Amount of money paid is a rough sort of estimate on some of the things we do, because the tendency is for people doing GROW and USE activities to get paid more than people doing largely WASTE activities. But this only holds as a very rough sort of rule. And there are plenty of blaring counter-examples.

And it collapses entirely when your governing apparatus has a lot of say in resource allocation and is focused obsessively on maximizing the short-term HWE number. An easy way to increase your HWE is to accumulate a lot of resources in exchange for future considerations(resources that probably would have been used largely on USE or GROW.) And then give it to people doing make-work or other tasks of dubious value (funneling lots of it straight into WASTE.) A secondary effect is to entice some people into spending some of their time and effort competing for the handouts (almost 100% pure WASTE.)

There are plenty of other tricks in the HWE-centric macroeconomic playbook. They will often have the advertised effect of making the HWE look good in the short-term, but have poorly-understood, unintentional, and generally undesirable effects on the mix of GROW, USE, and WASTE. And because the decision-makers get rewarded based SOLELY on short-term results, GROW takes a back-seat.

This is the fatal problem I have with macroeconomics as it is generally taught, discussed, debated, and practiced. It is a numbers racket, designed to manipulate a set of measures that don’t necessarily have much connection with the underlying reality they pretend to model.

"We're all dead in the long-run." - J.M. Keynes

[ I welcome all civil criticisms and comments, even sharply-pointed ones ]

(8 comments | Leave a comment)

September 25th, 2008
09:58 pm

[Link]

Voting Analysis, redux
candid told me I should have done the voting data analysis as a logistic regression, which was an excellent point. So I spent some of my "copious free time" re-learning how that works, and then re-parsed the NES survey data.

Population: Everyone in the NES survey who gave a valid responsive answer to every question pertaining to political affiliation and demographics data. I then culled out everyone who expressed no preference between the two major parties. N=823

The model tries to estimate P(R), the probability that a given voter in this population expressed a preference, to whatever degree, for the Republican party versus the Democratic party.

Variables found to be highly significant, and their weights:

INCOME +1.1 (this is a continuous variable from 1-23. The 1.1 represents the delta between bottom and top)
NHNCP -0.6 (non-hispanic with at least one parent who is not a U.S. citizen)
AGE<25 -0.6
POSTGRAD -0.8 (has a master's or PhD. MDs and lawyers did not fit the pattern)
OWNHOME +0.5 (indicated that they owned their home rather than rented)
WORKCLAS -0.5 (when asked if they self-identified as working class or middle class, chose the former)
WOMAN -0.8
RURAL +0.6 (from the possibilities rural,small town,suburb,big city,inner city)
SUBURB +0.7
PRAY12 +0.5 (when asked how often they prayed, chose one of the top 2 responses)
BIBLE1 +0.4 (believes the Bible is the literal word of God)
BIBLE3 -1.1 (believes the Bible was just written by men)
CATHOLIC -0.7 (self-identified as Catholic)
JEWISH -1.1 (self-identified as Jewish)
HISP -0.8 (self-identified as Hispanic)
BLACK -3.4 (self-identified as Black)

To put these numbers into some perspective, -3.4 is a GIGANTIC factor. All other variables being equal, it is enough to take someone from a 50% probability to 3%. A -0.7 swing, all things equal, changes someone from 50% to 33%

Variables that I tried in various forms, but weren't significant: Old age, being married, having kids, military service, unemployment status, retired status, treating income in any way other than a continuous scale, education other than POSTGRAD, any other way of considering urbanicity, religiosity, citizenship, or ethnicity

(Leave a comment)

September 15th, 2008
12:54 pm

[Link]

Six States
Obama : 255 base
McCain: 227 base

Ohio: 20
Virginia: 13
Colorado: 9
New Mexico: 5
Nevada: 5
New Hampshire: 4

The only other thing worth noting from my number-crunching the the noticable "corn-shift." Normally, Indiana would be unassailable R territory, and Iowa would be swing. Instead, IN has softened to merely strong R, and Iowa is semi-strong D. I can find no explanation other than corn.

(1 comment | Leave a comment)

September 10th, 2008
12:28 pm

[Link]

Red Voter, Blue Voter
The American National Election Studies is a collaboration between the U of Michigan and Stanford. They do polling of individuals during election years, asking them various questions about their political views, views of candidates, and demographics data. They make the full dataset available for download, so one can do data analysis.

N=939 (everyone who gave valid answers to the pertinent questions)

Independant variables:
B1: gave answer #1 to the bible origin question (roughly, the Bible is the word of God) [0-1]
B3: gave answer #3 to the bible origin question (roughly, the Bible was just written by men) [0-1]
M: answered "yes" to married or attached [0-1]
P: has a post-grad degree [0-1]
A: Self-identified as african-american [0-1]
H: Self-identified as hispanic [0-1]
I: Answer to household income level, on a 23-point scale [1-23]

Dependant variables:
DT: Bush temp + Rep party temp - Kerry temp - Dem party temp (temp is a "temperature" gauge from 0-100 on how much the respondent likes the entity)

Adjusted R-squared = .17

Intercept = -26.9
B1 coefficient = 21.7
B3 coefficient = -51.6
M coefficient = 19.6
P coefficient = -25.5
A coefficient = -76.1
H coefficient = -26.0
I coefficient = 1.92

Notes:
All coefficients were significant to 99%
No other way of chopping up education produced 95% significant results
No other ethnicity responses were 95% significant
No other way of chopping up income produced better results than simple linear
The questions asking about military service were not 95% significant
I failed to include any variables for children in the household

There's lots more that can be done with this dataset. This was just the result of me goofing around with the data for a couple of hours on Saturday night.

(3 comments | Leave a comment)

June 14th, 2008
09:55 am

[Link]

Future Economist
Couple nights ago, I was kid-watching so Sister and Wife could have a "wild night" out on the town. Four kids is usually easier than two, because with two, I am mainly the entertainment. With four, I am occasional referee.

Anyhow, we were winding down with some relaxing Unbeatable Banzuke on the TV, and when the show was over, it was time to send them off to bed.

Me: "Alright! Everyone head on out! Make sure you have everything you need for sleeptime!"
DtY: "You mean WANT, Daddy."

(2 comments | Leave a comment)

June 9th, 2008
04:57 pm

[Link]

Just up to our tip-toes
My sister is visiting this week, so we also get Niece and Nephew, about the same age as the Daughters. Thick as thieves, the four of them.

I was allowed to sleep in this morning (meaning 8:00 wakeup), and headed out to find Wife and Sister hanging out in the living room. The house was quiet.

"Where are the kids?"
"I dunno. Outside."

Fair enough. So, I mosey along to the front door, and poke my head outside. I hear kid noises, so I wander out to the driveway to see what's up.

I immediately notice Daughter the Younger, looking excited, wearing her helmet. Daughter the Elder is helping Nephew with the other helmet. I just sort of assume that they are going to ride on scooters.

But Daughter the Younger has a terrible poker-face, and is exuding a strong "Oh Noes!" vibe.

Me: "What's up, buttercup?"
DtY: "Nothing."

Me: "Look, I just want to know what you guys are doing."

*awkward pause*

DtY:"We're opening the garage doors, and you hold on to the handle. To go up."

*blink*
*blink*

DtY:"But just up to our tip-toes!"


At least they were wearing helmets.

(3 comments | Leave a comment)

April 19th, 2008
10:46 pm

[Link]

Valhalla Saturday
Thursday was an scheduled off day, and I missed Friday.
I did some "girl carry sprints" to try to atone for the miss:

Pick up 80lb girl in a fireman carry. Sprint 40m. Race girl back to start.
Pick up 55lb girl in a fireman carry. Sprint 40m. Race girl back to start.
30 seconds rest.
Repeat.

Saturday, the girls and I went bike riding some before bed. Headed down to the park where they helped me do sort of a make-up for Tuesday.

200m dash (actually 70-pace section back and forth 3 lengths)
10 push-ups
rest 1 minute

2 reps. For time, clock stops during rest.

Daughter the Elder reported 66 seconds and 92 seconds for the two sets.
I calibrated her counting at home against an actual clock, and she was within 10%, so we'll call that accurate.

Tomorrow I will tackle Saturday's workout.

(Leave a comment)

April 17th, 2008
12:49 am

[Link]

Valhalla Week, Wednesday
I missed Tuesday. My excuse is that I barely had enough time to slip away for "dinner" with the family, which consisted of snuggling a bit with the daughter who had an ear infection and coloring for 10 minutes with the one who did not.

Tonight's dinner break with the family was much more upbeat, what with the ear infectoid now feeling better and bouncing off the walls. So after a quick dinner (shredded chicken, cabbage, onion, cumin for me), I enlisted the daughters to help with Wednesday's exercise.

4 sets of
10 burpees each
1 minute rest between sets, clock stops during rest
For time
1:59*

*I don't think I did them quite right. Standing-squat straight down-kick legs fully backward-bring legs back under body-stand. I guess I was supposed to do a pushup there while my legs were out.

The Daughters wanted to do it, too, so they banged out their even-less-strict burpees while I was in rest period.

I think the reduced difficulty burpee technique was for the best, because ohhhhhhh owwwwwwwwwwwww aaaaaaaarrrrrrrrrr. Ended with me lying on the floor curled up and nauseous and sucking wind. Daughter the Younger came over to look down at me, then returned shortly with a paper towel. "You're all sweaty." Thanks, kid.

Oh, and I was wrong about not getting sore after Monday's workout. Nothing Tuesday, but today I hurt in a bunch of little places throughout the hip and abdomen. Places I'm not used to aching.

(2 comments | Leave a comment)

April 14th, 2008
05:06 pm

[Link]

Valhalla Week, Monday
This week, I'm following the exercise regimen spelled out here:
http://www.vcrossfit.com/200804/one-week-plan/

Today I did
12-9-6 (reps; 12 reps first round, 9 reps second round…)

dumbbell thrusters (Using 15 lb dumbells)
leg lifts (on my back, lifting knees to my chest)
push-ups (normal pushups)

For Time. (clock running throughout, taking breaks is OK, but its on the clock)

6:00 total. Each round, the thruster took up around 30-40 seconds, the leg lifts went fast, and then the pushups took about a minute, with 1-2 breaks. Then 10-20 seconds of recovery and back in.

No muscles feel like they were "worked out," but the system as a whole was certainly red-lining. Quick cardio ass-kicking. Yikes, I'm already dreading Saturday.

(3 comments | Leave a comment)

[<< Previous 10 entries]

Powered by LiveJournal.com