Tuesday, November 27, 2007

A spreadsheet to model the economy.

Back in January of 2005 I spent a week (in January, during my winter vacation) researching public spending at the federal, state, and local levels in the United States. I gathered all the numbers I could from various government web sites and the research papers available from think tanks and information clearing houses. I took all this information and used it to think through some of the problems of our economy with the use of an Excel spreadsheet.

In this spreadsheet I tried to create a rough approximation of the American economy, and then invited myself to figure out how I would set taxes and allocate public spending to meet the needs of my society.

I’ve shared this spreadsheet with some students, and now I’ll share it here on this blog. Let me give some explanation of the various worksheets and things in this spreadsheet.

Download the Spreadsheet in Microsoft Excel format by using this link here. If you are using iWork'08 instead of Excel you can still download that file and edit it.

There are seven worksheets.

The first worksheet is called SET UP QUESTIONS. A orange band runs down the center of this worksheet. On the left side of this band are questions and variables where you the user may change dollar amounts and percentages to set up your taxation schemes and spending allowances exactly as you would like to. On the right side of the orange bar are some facts about the economy modeled in this spreadsheet. There are also some facts about America and American public sector spending from 2004 and 2005.

The second worksheet is called COMPARISONS.
This compares the total dollars you’ve allocated to various public goods and policies and shows what percentage of the GDP you’ve put into these things. Then, to the right of that I have tried to get the most accurate figures I could for total public spending (all Federal, State, and Local government spending, not double-counting grants to states or local governments given by federal or state governments). These figures are as accurate as I could make them with just a few days of research, but I found figures that seemed to be fairly accurate. My figures differ slightly from some authorities. In 2004 it seems to me that I found indications that all public spending on education was about 5.3% of the US GDP, but an authoritative source in 2007 claimed we spent 4.8% of our GDP on education. I added the spending on military pensions, veterans hospitals and benefits, and those sorts of budget items to the defense for a total military budget rather than putting those pensions and health care spending categories in with things like health coverage, old age pensions, or disability benefits. My figure ended up being about 4.35% of GDP on defense spending, which ought to be a little high because I included some things that aren’t usually included, but in fact I found some authoritative sources claiming the USA spends 4.5% of its GDP on defense, so I’m not sure how people arrive at that higher figure. But, where I was able to check my final figures with those from other experts, it seemed my figures were quite close to what others had reached, and I’m satisfied that my numbers are approximately correct.

A huge part of state government spending is put into the pensions for retired state workers. Many states cover this by withholding a portion of state employee pay and using these funds to pay state employee pensions. Since the withholding from state employee salaries isn’t really “tax revenue” it seemed unfair to count state worker pensions as a sort of public spending based on tax revenue. Nevertheless, I think I stuck the state worker pension estimates in with miscellaneous aspects of public spending or administrative costs of running the government, and I did not include those in the category of “old age pensions.” If I were doing this again I would have put state worker pensions in with the social security payments.

Anyway, that comparisons page has a column where you can compare how your spending compares to actual public spending. When you get the spreadsheet as I have set it up the total military spending is about $124 billion, or a little over 1% of our GDP, this is about a quarter (or less) of what our government actually spends on military defense and associated costs, so in the comparison table (as it exists before you change anything) the “difference” number for military spending is -75.7%. That’s a pretty steep cut.

The third worksheet is WELFARE CALCULATIONS.
This page takes stuff given to the simulation on the page of set up questions and calculates how much the old age pensions and education system will cost. There are also some facts I found about labor force participation and food budgets. Some cells on this page help calculate the food supplement payments as well.

The fourth worksheet is INCOME DISTIBUTION CHARTS
It’s really just a single chart showing the income distribution of the United States in 2003 or 2004 and comparing it to the income distribution of the model I’ve made. My model has more poor people (actually quite a few more) and fewer super-rich. So, the economy model I’m using with this simulation is not really as wealthy as the United States. But, I think my model is pretty close to what we have.

This is a chart where the x-axis is a person’s income before taxes and benefits, and the y-axis is a person’s take-home pay after taxes and benefits. You can see that in the model you get (before you change any of the variables) there is this great leveling effect where as earnings go up from $30,000 per year to $85,000 per year (for able-bodied working-aged persons) actual take-home income only increases from $30,000 to $43,300 (after taxes and benefits). This means that people take home only about 24 cents for every dollar raise they earn between the incomes of $30,000 and $85,000, but once they get to the maximum tax rate at $85,000 (of 50% in the opening settings) they start keeping 50 cents for every dollar increase they enjoy in wages. It’s just a mathematical fact that if people at the lower tax rates get more in benefits and income than do people in the middle tax rates, and if there is a flat tax at some highest level of taxation, then at some point in the income distribution between the lowest rates and the highest rates people will be paying more than the highest rate on every additional dollar they earn. Someone must pay a higher marginal rate of taxes on their increases of income as the tax rates increase from zero to some highest point. See if you can play with the model and get around this. You can’t. It’s impossible.

The sixth worksheet is INCOME DISTRIBUTION.
This is the model economy. I used this to make the chart in the income distribution charts.

The seventh worksheet is POPULATION INCOME MODEL
This is the sheet where I model the whole population in terms of their earnings and their benefits.

Some more about this model.

In the real American economy the "public cycles of money" collect revenue from you in a wide variety of ways. You may pay property taxes, registration fees, user fees, sales taxes, income taxes, and various special taxes on specific products. These taxes are imposed by local, state, and federal governments. It’s difficult to keep track of all the various taxes you’re paying. In this simulation, we’re saying for the sake of simplicity that the government has abolished all sales taxes and property taxes aside from a few taxes on luxury items or items that injure health. Second houses, vacation houses, and boats might be taxed. There still might be taxes on gasoline, cigarettes, alcohol, gambling winnings, and that sort of thing. But in general, for most purchases, sales taxes have been abolished. The only main tax that people must pay are income taxes. And all incomes taxes are rolled into one big income tax, from which revenue is divided among the local, state, and federal government. So, the only tax you need to set is this one income tax. Currently people in the highest tax brackets might pay something like 35% of their income in federal income taxes, 3% of their income in state taxes, the equivalent of 4% of their income in property taxes, and the equivalent of 5% of their income on sales taxes and special fees for government services, toll roads, and so forth. Imagine rolling all those taxes into a single tax rate of 46% (this is assuming that about 1% of their income still went to the government in luxury taxes on their second houses or gasoline taxes, and that's included in the 46%). Talking about a 45% or 46% income tax to someone who is accustomed to paying only 35% of their income in federal income taxes might seem like a terrible tax increase, but in the case I’ve outlined above this would be a revenue-neutral shift of all tax burdens into one simple income tax, and the rise from 35% to 46% in income tax would be matched by a drop in property and sales taxes.

How to play with the numbers

The model starts with the big ticket items. After you have set your tax rates and your Earned Income Tax Credit (a sort of negative income tax that gives money to people who earn very little instead of taxing them, but gives more money to people who earn some income and less money to people who earn nothing), you get to decide about issues like retirement age, size of retirement pensions, disability incomes, national health insurance, education spending, unemployment insurance, housing allowances, food spending allowances, and defense spending. When all these are done, you are told how much money you have left, and you allocate it among the remaining categories by deciding what percentage of the remaining money should go toward various things.

Be sure not to short-change your public safety and justice spending or your basic infrastructure spending. Public spending on courts, police, fire protection, and prisons in the USA was over 200 billion in 2004, and you probably don’t want to cut spending too much here (although legalizing marijuana and taxing it would cut some of these costs and raise revenues, but I don’t know how much that would actually cut law enforcement expenses.) Spending on infrastructure was also just above 200 billion. If you pay much less than this for infrastructure your bridges will start collapsing, your roads will become full of potholes, and sewage treatment and water treatment will become unsafe. Don’t short-change enforcement of regulations, either. You want to ensure people aren’t eating feces and bacteria when they buy their food. If you don’t enforce workplace safety, food safety, or drug and medical standards you are going to have some angry people kicking your head when their children die from eating rotten hamburgers or tainted cough medicine.

See if you can balance the budget, figure out a “fair” system of taxation, and eliminate poverty!

For the sake of simplicity, everything in this model assumes a very simple welfare system. Everyone gets the same old-age pension, no matter how much you paid in taxes. That’s not the way it works now. For health benefits, there is one standard health plan that covers everything you can cover based upon your national health spending budget. A certain amount is allocated for the nation’s health care needs based upon the average cost per person for all health care (at least all health care covered by the national health care plan. Maybe in your fantasy health care plan something isn’t covered). Public housing has been abolished and the government just gives poor people a “housing allowance” (money that can only be used to pay for housing and utilities). A similar system exists for food security, with every person at a certain income level getting food money (like food stamps) that can only be used for food. If you’re not disabled, not unemployed, and younger than retirement age, then you get an EITC to supplement your income plus the housing money and the food money. That’s it. There’s nothing like TANF or AFDC.

In this model every child is given an education voucher equally to 100% of the costs you’ve set for various levels of education. A public school is any school that offers an education that has total costs equal-to-or-less-than the costs you’ve set for education, and does so while obeying the constitution (no establishment of religion, so no forcing any particular religious beliefs on students, and no discrimination in violation of the equal protections amendments). Any school that requires any money in excess of the officially set cost of school is by definition a private school. Any student enrolled in such a school does not get to use a voucher at that school. School vouchers are only good at public schools. Public schools are welcome to spend more on their students than they raise through vouchers (if they raise additional resources through alumni giving, donations by benefactors, and so forth). A national set price for public education at each level would create a powerful incentive for schools to keep costs down to the (hopefully reasonable or generous) levels you have set. If schools allow costs to exceed what they bring in through vouchers they can’t pass on the extra costs to the students and their families, nor can they coerce local residents to pay for the increased educational costs through broad-based local property or sales taxes.

The assumptions and simplifications limit the accuracy of this simulation, but it’s still a fun game. Go ahead and download the spreadsheet and have some fun pretending you’re the president and the legislature all rolled into one. I’d be interested in knowing whether this game gives you any new insights into how taxing and spending really works.

Thursday, November 15, 2007


As an educator in a higher education setting, I'm interested in the purposes of higher education. What should our university's students be getting out of their education? Ernest Boyer (1928-1995) provided many of the ideas that guide my university. He once said in an interview shortly before he died, that:

. . . Every one of our Founding Fathers knew that if we wanted to move toward a government that was run by the people, they had to be enlightened. Surely, they have to work; surely, they have to be responsible as producers as well as consumers. But the larger purpose of education in this country is always driven by the fact that we need people to be civically engaged, intellectually and educationally well informed, or else we were opening the doors to tyranny. . . .

I agree with that. I think that even in a university setting we professors and staff must fulfill a duty to model a civically engaged life, and we must be certain that our students graduate with a fair understanding of some of the basic facts, problems, and theories related to government, history, and economics. In fact, I've been trying to work with some of my colleagues in our department to come up with some ideas about core knowledge that we agree our students should know, because I would like to measure how much of this core knowledge our students know when they enter our program and how much they know when they graduate.

So, it was with great interest that I took a civics test an old friend, Scott Fares had posted on his blog. This test is the civics literacy test produced by the Intercollegiate Studies Institute's National Civic Literacy board. Checking out the mission and values of the Intercollegiate Studies Institute (which you can perhaps guess as you take the civics literacy test) I found six values guiding ISI. I strongly agreed with two of these values: Individual Liberty and The Rule of Law, but I could only give partial or strongly qualified agreement to the other four values. In fact, two of the so-called "values" seem more like empirical (and falsifiable) propositions to me, rather than values (these are the so-called "values" of the Free Market Economy and Moral Norms. I think there is some truth in these propositions, but I wouldn’t entirely endorse them as true.

Anyway, I did pretty well on the test, scoring 57/60 for 95%, and upon reflection I realized I had known the correct answers to two of the three I missed, but I hadn’t taken the time to think carefully about what the questions were asking and I hadn’t bothered to read all the potential answers carefully. One item on the test (number 58, What is a major effect of a purchase of bonds by the Federal Reserve?) completely stumped me.

Today in my liberal studies class I had my students take the quiz. Scores averaged 31.5 correct (52.5%), which puts my students, who are mostly first-semester juniors, although I have some first semester seniors and second-semester sophomores, in the company of seniors at schools such as Iowa State University, the University of Montana, the University of Florida, Gonzaga University (in Washington), and the University of Michigan.

As my students took the test, I re-took it, but instead of looking at the quiz to consider the correct answers, I tried to rate the quality of the questions. Many of the questions, especially in the history section, seemed to me to be asking questions that were not especially significant or important. For example, the first question asked students to identify the century when the Jamestown colony was founded. Well, it was founded exactly 400 years ago this year, so people really ought to know it was founded in the 1600s, but if someone mistakenly confused Jamestown with Roanoke (which was founded and disappeared late in the 1500s), would that really indicate a woefully inadequate understanding of American history? I think not. Now, someone who thought Jamestown was founded in the 14th century would be demonstrating a complete lack of understanding of chronology, but the test makes no allowances for distinguishing between close misses and really awful guesses.

I also found many brilliant questions. In fact, the good (and even great) questions outnumbered the weak questions. Only a very few questions were genuinely bad or misleading. All in all, I think I’ll use this quiz in my future classes and my senior seminars to see how our students score compared to the scores the ISI is getting at other schools. I could think up an additional 20 questions to mix in with the existing 60 and add those with the 40-or-so good questions in the existing survey to make a new alternative version.

One weakness of this sort of testing is that it mainly measures knowledge of facts, familiarity with terms, and frequently we’re just learning how well students can recall of details of dubious importance. It’s difficult to measure analysis, synthesis, comparison, evaluation, or application with short and simple multiple item tests, but there are great items on this quiz that do measure something beyond simple recall of facts, and I’m thankful to the ISI for making this quiz free and widely available by publishing it online.

Before taking the quiz I had my students watch an episode of the old 1960s television miniseries The Prisoner staring Patrick McGoohan. This was the episode called “ The General” (episode six). The episode features a science-fiction technology called “speedlearn” in which college courses are impressed directly into the cerebral cortex through television transmissions, so that students can memorize the facts of a course in just a few seconds of television viewing. There is a great scene where the authority figure is describing how this technology will enable experts to put infallible information into the minds of students with minimal effort, and asks “what have we got?” The Patrick McGoohan figure, who plays the hero, retorts with disgust, “a row of cabbages” to which the authority figure responds, “indeed, knowledgeable cabbages.” The Prisoner asks about what people should learn, and who chooses. A good question. I’m sure I don’t want the ISI to be in charge of deciding what information my students should be learning in college.

My students discussed this with me throughout the semester, and each one writes a paper on their educational philosophy. What should people learn? Who should decide what people must learn? Several students share the opinion held in common by the ISI and myself that there are some core knowledge facts that everyone with a college degree ought to know. Some of these facts are measured by this ISI quiz we took, but many crucial facts are missing, and many facts that the ISI thought important enough to include in the test seemed trivial to me. There are a few questions that are only generally, but not precisely correct (e.g., the battle of Yorktown effectively ended fighting in the American Revolution, but it didn’t end the war, which wasn’t concluded until the Treaty of Paris was signed in 1783, nearly two years after the Battle of Yorktown. On the other hand, The War of 1812 was concluded by the Battle of New Orleans, which was actually fought after the Treaty of Ghent had officially ended the war, but since the war of 1812 was in many ways a continuation of the American Revolution war, it would make sense, and be highly original and creative, to say that the Battle of New Orleans concluded the wars of American Independence, whereas the Battle of Yorktown merely initiated a 31 year pause in the active warfare between England and some of her North American colonies).

We discussed in class some of the issues that face us if we try to determine what is important to know in history or government or economics. I mentioned a lecture by James Loewen I had attended this past weekend on the subject of the Springfield Pogrom against African-Americans and the nadir of American race relations and the problem of sundown towns. In my experience, the sort of people that advocate for the sort of test the ISI produced seem to think it’s more important to know that Jamestown was settled in the 17th century (and not the 18th or 16th century) than it is to know that between 1891 and the 1940s America went through a nadir of race relations in which many small towns chased out Jews and African-Americans, and mobs in many larger towns tried to destroy African-American neighborhoods and terrorize African-Americans (there were pogroms in Tulsa, Chicago, Springfield (IL), East St. Louis, Wilmington (NC), Atlanta, and Detroit). I do think it’s important to know when Jamestown was settled, or when Lincoln was elected, or the basic chronology of the Civil War, but I think it’s significantly more important to know something about the history of race relations in our society, because this is something that still influences us.

Instead of asking about the years when Lincoln served as our president, I’d rather ask about the content and meaning of his second inaugural address. Instead of asking when Jamestown was founded, I’d rather ask a question about the importance of the plantation system (especially for the growing of tobacco) in early Virginia history and how this plantation system set up an economy where many laborers were imported in slavery or indentured servitude from the British Empire or from Africa, and how this made Virginia culturally and economically quite distinct from the New England colonies of small landholders.

If you take the quiz, let me know how you do, and what you think of the exam. Can you beat my students’ average score of 32 correct? Check out Loewen’s page on sundown towns. What do you think? Should this be part of core knowledge that all American college students learn? How does it rank in importance with other basic facts?