This one speaks for itself:

Files on abused children. Employee evaluations. Tax returns. A list of computer passwords. Names, addresses, birth dates and other information on hundreds of foster children and abused children. And, of course, Social Security numbers.

The information could hardly have been more sensitive — the raw material of identity theft and invasion of privacy — yet the State of New Jersey was about to turn it over to the highest bidder, the state comptroller, Matthew A. Boxer, reported on Wednesday. After the comptroller’s office reviewed computer equipment that the state was preparing to auction to the public last year, it found that 46 out of 58 hard drives, or 79 percent, still had data on them, much of it confidential.

The report from the Comptrollers office can be found here (pdf).

Speaker Quinn announced on Monday that a living wage bill will get a hearing in April. The bill would require buildings and developments that get city subsidies to pay higher wages to their workers. Although, in theory, wage floors can cause disruptive effects on the labor market, I think bills like this make good sense. The city is heavily subsidizing some developments through tax breaks, and it makes sense to demand that those developments treat their workers a little better than they otherwise might.

A lot of developments get their subsidies because they have affordable housing. That works by giving developers large square footage bonuses for the amount of affordable housing they build. Maybe it would make sense to do something like that in the case of the living wage: a development can have 5% more square footage if there is 30-year agreement to pay the workers a living wage with a cost of living adjustment. I don’t know what percentage of square footage would make something like that work for most developments, but it would address the complaint that the bill will “drive up the city’s economic development costs and prevent new investment in city neighborhoods” while providing better wages to workers in the developments. It may seem counterintuitive to link land use questions with employment, but why not?

The Department of Health and Mental Hygiene issues 3,100 year round permits for $200 a piece. There is a long wait list (PDF). The Wall Street Journal had an article this morning on the rampant black market that exists for these permits:

Meanwhile, demand for permits and their black-market prices continue to climb as street food’s popularity soars with blogs like Midtown Lunch chronicling vendors’ moves and some gourmet food trucks developing cult-like followings. Some permits fetch as much as $20,000 for two years, vendors say. In the case of Ms. Sultana, the Bronx food vendor, she says the permit holder told her someone else was willing to pay $15,000 for the permit she previously paid $7,000 for two years ago.

Mohammed Rahman, who has operated the popular Kwik Meal cart in midtown for 11 years, says he pays $15,000 every two years for his permit. “The city charges only $200, why should I have to pay $15,000? All the profits go to someone else.”

My first reaction upon reading this article is that there should obviously be a legal market for street vendors run by the city. The article talks about how there could be “big money” in the City if they charged more for the licenses, and I agree. I don’t think the permits should be set artificially high, but maybe there should be an auction for food cart permits (I suggest checking out some of the work done by Susan Athey). You could imagine bids for food cart permits in such an auctions, and then charging everyone the price they bid (we tend to have some concerns with price discrimination in the public sector), and then everyone pays the lowest big. That’s a pretty good system actually, since given the wait list there are more than 3,100 people who want the permits there would be more than 3,100 bids, but the price is likely to get too high. It is also difficult to imagine massive collusion, since everyone has an incentive to bid more than they think the lowest bid would be (lest they accidentally are on the wrong side). The most profitable street vendors with most certainly win, which is a good characteristic, but there would be plenty of room for mid-range food carts.

This is the first in a series of post where I am converting lecture notes and class slideshows from P11.2140: Public Economics and Finance in to blog posts, mostly for my understanding. The class is taught by Nirupama Rao, and some of her lecture slides and diagrams come from the book Public Finance and Public Policy by Jonathan Gruber. The class uses no calculus, which is disappointing to me, but it is pretty straightforward to understand all the math.

Public finance as an academic field is the study of the role of government in the economy. A larger part of it is concerned about taxation and subsidies, although the provision of public goods is definitely within the bounds of public finance. Public finance is considered an applied branch of microeconomics.

Public finance seeks to answer four main questions according to Rao/Gruber:

When should the government intervene in the economy?
How might the government intervene in the economy?
Why do governments choose to intervene in the ways they do?
What are the effects of alternative interventions on economic outcomes?

Lets take each of them up in turn.

When should the government intervene in the economy?
The two cases when a government should intervene in the economy are market failures and redistribution.

Market failures occur when, well, the market fails at the optimal provision of some good. The first classic examples of market failures are externalities, which are when there are “spillover” benefits or costs on to others. In the case of something like pollution, I’ll over produce my good because I don’t bear the full cost of it. A positive example is vaccinations, where individuals might under purchase/produce them because the private benefit is smaller than the public benefit of everyone being vaccinated.

The second market failure is the provision of public goods. Public goods are defined as non-rival and non-exclusive. Non-rival goods are ones where my consumption doesn’t effect other people’s consumption of that good. Non-excludable goods are those that you can’t prevent people from accessing it, even if you wanted to. Examples of public goods that come up a lot are fireworks and national defense.

The third market failure is imperfect competition. The purest cases are when there is one seller (monopoly) or one buyer (monopsony).

The fourth and last common market failure is asymmetric information, which is when one party in a transaction has more information than the other. The most cited case is adverse selection, and the most famous paper on that topic is The Market for Lemons: Quality Uncertainty and the Market Mechanism (PDF).

Redistribution is much simpler to explain: the government often wants to shift resources from one group in society to other groups.

How might the government intervene in the economy?
The three ways that government can intervene in the economy is through prices (tax or subsidize), regulation (mandate or restrict), or provision (either through financing or direct provision).

Why do governments choose to intervene in the ways they do?
We basically don’t talk about this question in my public finance class. This is really the purview of political economics. Political economics often has two distinct meanings: “Leftist economists, suspicious of the market, called themselves political economists to emphasize that economic phenomena were largely politically determined. At the same time, more conservative economists who were enamored of the market called themselves political economists when they applied the tools of market analysis to political phenomena” such as how the government makes decisions (The quotation is from “Political Economy in Political Science” an unpublished note by Andy Eggers). We mean political economy in this second sense.

What are the effects of alternative interventions on economic outcomes?
The effects are usually split in to two groups: direct and indirect effects. Direct effects are the effect of a government intervention if behavior stayed constant. For example, if we lowered the tax rate, we would get X fewer dollars if every behaved the same. Indirect effects are the effects of a change in behavior in response to a government intervention.

This week’s On Education column is about the usefulness of some measures in ascertaining teacher quality. The article tells the story of an obviously dedicated teacher, Stacey Isaacson, who is ranked “7th percentile among her teaching peers — meaning 93 per cent are better” despite having “65 of 66 scored proficient on the state language arts test.” The article reflects some healthy frustration about testing, test scores, and what exactly value-added means, but overall I found the article very annoying. Let me take a segment of the article:

Everyone who teaches math or English has received a teacher data report. On the surface the report seems straightforward. Ms. Isaacson’s students had a prior proficiency score of 3.57. Her students were predicted to get a 3.69 — based on the scores of comparable students around the city. Her students actually scored 3.63. So Ms. Isaacson’s value added is 3.63-3.69.

What you would think this means is that Ms. Isaacson’s students averaged 3.57 on the test the year before; they were predicted to average 3.69 this year; they actually averaged 3.63, giving her a value added of 0.06 below zero.

Wrong.

These are not averages. For example, the department defines Ms. Isaacson’s 3.57 prior proficiency as “the average prior year proficiency rating of the students who contribute to a teacher’s value added score.”

Right.

The calculation for Ms. Isaacson’s 3.69 predicted score is even more daunting. It is based on 32 variables — including whether a student was “retained in grade before pretest year” and whether a student is “new to city in pretest or post-test year.”

Those 32 variables are plugged into a statistical model that looks like one of those equations that in “Good Will Hunting” only Matt Damon was capable of solving.

The process appears transparent, but it is clear as mud, even for smart lay people like teachers, principals and — I hesitate to say this — journalists.

The last line is the entire problem with this section. It is his job to tell us what he’s talking about in a way we can understand. That is at the core of journalism. Instead he just seems lazy.

For example: “These are not averages. For example, the department defines Ms. Isaacson’s 3.57 prior proficiency as “the average prior year proficiency rating of the students who contribute to a teacher’s value added score.” It is an average. I don’t know how you can say something is not an average and then quote its definition as beginning with the phrase “the average.” Now that whole sentence from DOE is slightly confusing, but I think all it is saying is that 3.57 is the average prior proficiency of all those student who are in the value added model for Isaacson’s effectiveness (important to note here, its not her prior proficiency but that of her students). There are all sorts of reasons that the particular students that she had in her classroom might not be in the model. That is, some students are likely excluded from the model because DOE doesn’t have values for all the other 32 variables or the student has a documented learning disability.

Value added models are statistically complex – they usually have fixed effects, random effects, shrunken estimates, and much more. They are beyond the statistical expertise of most teachers, administrators, and – yes – journalists. But that isn’t the point. It doesn’t matter that it is a complex model. If that mattered we would be outraged at the fraud detection on our credit cards. No, the real criticism is that value added models are notoriously controversial and, in practice, not very accurate. That’s the attack that should be leveled. As RAND reseachers noted in a often cited review of value added models in education from 2004:

The research base is currently insufficient to support the use of [Value Added Methodology] VAM for high-stakes decisions. We have identified numerous possible sources of error in teacher effects and any attempt to use VAM estimates for high-stakes decisions must be informed by an understanding of these potential errors. However, it is not clear that VAM estimates would be more harmful than the alternative methods currently being used for test-based accountability. At present, it is most important for policymakers, practitioners, and VAM researchers to work together, so that research is informed by the practical needs and constraints facing users of VAM and implementation of the models is informed by an understanding of what inferences and decisions the research currently supports.

As stands, my first reading was that statistical model that are multivariate regressions with 32 variables are something on the order of “Good Will Hunting,” which is a little extreme. He should note that its not the models complexity but its usefulness that is at stake. Either way, instead of stating a useful point – that DOE is using a limited model in a overly serious fashion or value added models are contentious in the academic education literature – he attacks complexity for complexity’s stake.

The article does note the ridiculously large margin of error on these teacher reports: “Moreover, as the city indicates on the data reports, there is a large margin of error. So Ms. Isaacson’s 7th percentile could actually be as low as zero or as high as the 52nd percentile — a score that could have earned her tenure.” If your margin of error is that big, I would agree that you shouldn’t be making tenure decisions based on it.

More generally, I think he misses the point that Isaacson could be a bad teacher despite the glowing anecdotes. She could work really hard but teach very smart students and not teach them all that much year-over-year. We don’t know, that’s true, because the DOE is using a model with a lot of kinks and a crazy margin of error, but that is different than knowing she’s a good teacher, which is what the article suggests.

The New York Times had an article yesterday on the hot Muni Market: Opportunity in a Muni Maelstrom. People are getting out of the bond market, which has led to some good deals:

“Some of the trades in this market have been very sweet lately,” said Matt Fabian, managing director of the research firm Municipal Market Advisors. Muni yields began rising in November, and prices, which move in the opposite direction, fell, hitting a trough on Jan. 14, he said.

Swept up in the price movements were some red-hot munis — including those Cornell and Harvard bonds. (Although state and local governments, as well as water systems and sewer districts, are classic issuers of munis, universities may also offer them through various means.)Traders who bought a 30-year Cornell 5 percent bond in mid-January and sold it last week would have pocketed a quick 9.3 percent profit, he said. Trading the equivalent Harvard bond over the same short period would have produced a 6.3 percent gain.

This comes on the heels of Nouriel Roubini predicting $100 billion in muni defaults over the next five years and, as mentioned in the Broke Town, U.S.A. article, Meredith Whitney predicted hundreds of billions in defaults. Both of those numbers are likely wildly overstated (although I’m no municipal bonds expert), as notes: “Since 1930, not a single American state has defaulted, the report said, while corporate bonds have defaulted at an average annual rate of 2 percent.”

Equally safe, likely are bonds from stable institutions with clear revenue. One could have made a good amount of money last week:

Swept up in the price movements were some red-hot munis — including those Cornell and Harvard bonds. (Although state and local governments, as well as water systems and sewer districts, are classic issuers of munis, universities may also offer them through various means.)Traders who bought a 30-year Cornell 5 percent bond in mid-January and sold it last week would have pocketed a quick 9.3 percent profit, he said. Trading the equivalent Harvard bond over the same short period would have produced a 6.3 percent gain.

“Individual investors looked at the pricing and the yields of these bonds, and came to their own conclusion: a place like Cornell is going to be good for the money, and when the smoke clears, this will be a real opportunity,” said Thomas McLoughlin, managing director and head of municipal research at UBS.

The Mayor’s Management Report (MMR) is an bi-annual compendium that measures the delivery of certain things in every City Agency.  It is mandated by the Charter and comes out once a year. The Gotham Gazette reported this morning that the preliminary 2011 version has come out:

With very little fanfare — and seemingly not even a press release — Mayor Michael Bloomberg has released the Preliminary Fiscal 2011 Mayor’s Management Report, which provides an array of statistics on the performance of the city government for the first months of this fiscal year. The report — along with a full-year report released in the fall — is required under the City Charter.

So what’s in the MMR? Some examples I picked randomly:

• The number of people who smoke in the City stayed steady from FY09 to FY10 at 15.8%, but the FY11 target is 13.2%.
• Percent of Cash Assistance applicants and recipients placed into jobs as compared to monthly goal was 99% in FY10, but is over 100% in the last four months.
• Although there is a target of 100% of abuse and/or neglect reports responded to within 24 hours of receipt from the State Central Registry, the actual rate was 95.8% in FY10.
• It took 16.1 days for the Department of Environmental Protection to fix most sewer leaks.
• The city received 31,778 noise complaints in FY10.
• 96.1% of all streets were rated acceptably clear (whatever that means) in the last four months by the Department of Sanitation.
• 94% of the Department of Parks and Recreation’s Water Fountains were working last year (not my anecdotal experience in Central Park).
• The were 37,597 major felony crime this quarter.
• 63.0% of parking tickets are paid within 90 days.
• 757,501 people attended programs at the NYPL in FY10.

And many more interesting statistics. If you want to learn about a City agency, one of the best ways to do it is to go through the few pages on it in the MMR.

For reference, Fiscal Years end in the number they’re labeled.  NYC, oddly, has a fiscal year of July through June. So FY11 is July 2010 through June 2011.

I wanted to briefly explore Thomas Piketty and Emmanuel Saez’s well known 2007 article from the Journal of Economic Perspectives called How Progressive is the U.S. Federal Tax System? A Historical and International Perspective. Their basic finding is that the Federal Tax system has become less progressive over time.

They find that average federal income tax rates for the middle class have stayed the same from 1960-2004, but that a drop in corporate taxes “combined with a sharp change in the composition of top incomes away from capital income and toward labor income” has led to the drop in the system’s progressivity (pg 23). They also point out — this is crucial for debates about taxes — that the marginal income tax rates haven’t been where the rich have gotten the most gains. Rather, it is in the deductions and exemptions that have been created. As might be expected:

Large reductions in tax progressivity since the 1960s took place primarily during two periods: the Reagan presidency in the 1980s and the Bush administration in the early 2000s. The only significant increase in tax progressivity since 1960 took place in the early 1990s during the first Clinton administration (pg 23).

You can see this change in the mix of taxes that people face (note the large change for the highest percentiles):

We can also see in the chart below that the share of total income for the top 0.1% have gone up over time, as the average tax rate has decreased and the share of total taxes has increased:

They note in their conclusion:

[T]he most dramatic changes in federal tax system progressivity almost always take place within the top 1 percent of income earners, with relatively small changes occurring below the top percentile. For example, many of the recent tax provisions that are currently hotly debated in Congress, such as whether there should be a permanent reduction in tax rates for capital gains and dividends, or whether the estate tax should be repealed, affect primarily the top percentile of the distribution—or even just an upper slice of the top percentile. This pattern strongly suggests that, in contrast to the standard political economy model, the progressivity of the current tax system is not being shaped by the self-interest of the median voter (pg 23).

We can see that happening right now in New York, as we are very focused on the “Millionaires Tax,” which seems like a common sense progressive tax proposal given the state’s dire fiscal condition.

“Determining who will suffer from budget cuts is a political and a legal calculation.” That quotation from the middle of Broke Town, U.S.A., is the sentiment at the center of this blog. I have many friends who are passionately interested in public policy, but are rarely interested in the budget. It is a bunch of numbers in tables over hundreds of pages: How interesting could it be?

It is true that the tables can look boring. Earlier today, I was looking through the line-item departmental budgets from City Agencies and OMB, which clocks in at 3950 pages. It looks like it was printed in 1985. Yet it encompasses every service provided by the City. How many police officers are there? How much money are teachers given for classroom supplies? How much money goes to Special Education students? How many social workers are there for Child Protective Services? Can the City afford an extension of the 7 subway line to New Jersey? The answers to these questions are all in the budget.

Back to the article. It does a great job of synthesizing a lot of different topics. The overarching question is about municipal debt:

If muni bonds were to default (causing investors permanent harm, as distinct from the temporary discomfort of price fluctuations), ordinary Americans would lose big. Munis are bonds issued by state and local governments, as well as agencies like hospitals, with the interest going to bondholders tax-free. Their relative safety, plus the tax break, has made them a favorite among individual investors, who own about two-thirds of the total, either directly or via mutual funds.

But what if the burden of municipal woes falls elsewhere than on bondholders? Yes, cities and states have creditors. They also have citizens who rely on their services and who pay the taxes, and they have public employees who are dependent on stable public-sector jobs and often-ample benefits. Whitney isn’t wrong about a crisis in local government; the crisis is here. The question is, will it be articulated in terms of bond defaults or larger kindergarten classes — or no kindergarten classes at all? The efforts in Wisconsin and elsewhere to squash organized labor suggest that politicians are no longer so willing to protect public employees. Teachers and nurses are likely to suffer well in advance of investors.

The article is definitely worth reading, and gives a good overview of bankruptcy (is it, for example, even an option worth considering?) and pension liabilities.

&nbs

From the December 2010 Jobs Report of the Center for Urban Research:

There is a concept in revenue analysis called revenue capacity, which is defined as the maximum revenue that a government can sustainably generate. Once that point is reached, revenue raising actions like increasing tax rates do not result in more revenue. The theory is built around the idea of competition between states, and also between municipalities. That theory is obviously overwrought — after all, there are a lot of reasons to live in New York that have nothing to do with the tax rate, and the simpler models do not account for, e.g., political or legal restraints, or for the rules of thumb that ratings agencies use to rate government bonds. Nonetheless, it is useful to understand the basics of revenue capacity and revenue reserve.

In the simplest form, you look at revenue capacity, which is defined as the amount of revenue you can raise when compared to other states. For each level of per capita GDP there is some level of revenue that can be sustained by “the market” of states. To come up with thus number, we run a simple regression of per capita tax income against per capita real GDP by state. The following diagram does that with each small circle being a state and the trend line representing the single variable regression:

The trend line is average revenue capacity. If a state is below that line that suggests they have room to raise their tax rates and still be competitive. If a state is above the line (as New York is) it means that the tax rate is already above the amount that is competitive at the GDP per capita. Again, this is a very simplistic model, which does not account for industry mix, how much people are willing to pay for services, or political history; however, it does give us a picture of where New York State’s revenue is when compared to the other states.

I got the total tax data from the Census’s 2009 Annual Survey of State Government Tax Collections, the population data from the Census (XLS), and the per capita GDP at the Bureau of Economic Analysis (BEA).

There is a very simple measure for economic diversity called the location quotient. It very directly gives you the share of industrial sectors in a local economy and then compares that share to the share in some broader reference group, often the United States more generally.

To use it, we first need to think about what exactly an industrial sector is. Fortunately, this has already been done for us by the North American Industry Classifications System (NAICS), which was a joint project of the Federal Office of Management and Budget (OMB) and counterparts in the Mexican and Canadian governments. Sectors can be broken down into more specific subcategories, with longer the NAICS sector numbers denoting greater specificity. Two digit codes are very broad things like mining, education, health care, government, etc. Codes go all the way down to six digits. For example, Bennett Midland, the management consulting firm I work at, might be classified as:

54 Professional, Scientific, and Technical Services
541 Professional, Scientific, and Technical Services
5416 Management, Scientific, and Technical Consulting Services
54161 Management Consulting Services
541611 Administrative Management and General Management Consulting Services

Back to the location quotient. The location quotient is defined as the share of employment in industry i in a specific geographic area divided by the share of employment in industry i in some reference group.

The Bureau of Labor Statistics releases County Level Employment and Wage data every quarter, with a three-quarter lag. So, on January 11, 2011 they released the data for Second Quarter 2010, which is available here. Specifically, they release data for employment by industrial sector for the ten largest counties. Let’s look at New York, NY:

In this table we see the ten largest private economic sectors and government. Right off the bat, we can see what mix of jobs the United States has and New York has. 20% of all New York City employment is in professional and business services, 20% is in government, and 15% is in financial services, while almost no one works in natural resources and mining.

If the location quotient is greater than one, then the industry has a higher concentration in the county than compared to the reference group — in this case, the US as a whole. Thus, the closer the location quotients are to “1″ for all the industries the more closely the region matches US sector diversity. New York has many more jobs in information and financial services than most of the country, which makes good sense. We also have higher than average concentrations in professional and business services, and in government.

But what does this actually mean? Well, for one thing, economic diversity is is a buffer against economic shocks and recessions. So the more diverse a city, county or region is the more likely it will be to withstand the difficulties associated with downturns. Unusually high concentrations of employment in industries which contribute to the tax base can pose serious fiscal problems for state and local governments. We have seen this in New York City. Sometimes we enjoy great benefits to having so much of our economy in financial services (see $2 billion recently gained in the corporate tax), but other times we suffer more because of our reliance on one industry. Thus, our very high location quotient in financial (and to a lesser extent information) services increases our economic volatility, which makes long term fiscal planning much more difficult.