Since at least the time of Huey Long, who enacted a homestead exemption that essentially protected all homeowners in the state from property taxation, Louisiana has been known for a deep aversion to property taxes.

Of course, it’s also been known for colorful political traditions. And property taxes and politics have traditionally crossed paths in the assessor’s office.

Louisiana elects its assessors, and New Orleans elects seven of them, each responsible for several wards  a remnant of the late 18003, when a series of towns were annexed into one metropolis. While the city eventually settled on a single mayor, the extra assessors remained.

Though each assessor only sets property values fora relatively small sliver the city, that doesn’t mean they lack power. When A.J. Liebling was researching “The Earl of Louisiana,” his book about Huey’s brother, Earl, he was shocked to learn that one of New Orleans’ assessors was among the city’s chief potentates.

Things have changed since then, but not as much as one might expect. The assessors keep traditions that could fairly be described as quaint. Many records are kept on paper, and the assessors’ computer system is down-right Jurassic. Each year the tax roll is opened for a two-week period in August. By custom, many city residents visit the assessor during that time to make sure their properties are not being revalued. During the rest of the year, it’s always been difficult to get information about assessments.

Given that rich backdrop along with quite a bit of anecdotal evidence we had always figured a good story would come from analyzing assessment practices in New Orleans. Several factors converged more or less simultaneously to help make the story happen.

First, the new mayor claimed that, because of under-assessment of property, the city was losing $15 million to $20 million per year, an assertion the assessors hotly disputed.

The state’s legislative auditor issued a report that blasted the practices of 12 of the state’s 70 assessors, including one from New Orleans. The report said assessors were not changing the values on most properties everyfour years despite laws requiring a quadrennial reappraisal of all property.

And the mayor’s staff took the revolutionary step of posting tax rolls on the Internet. The rolls technically belong to the assessors, but city staffers run the assessors’ computer system. The assessors fumed, and even contemplated a lawsuit, but the damage was done. The veil had been lifted.

Our conundrum was how to come up with a method to analyze assessments. The key question seemed to be: Are assessments accurate? The only practical way we had to measure that was to compare assessments to sales prices, which are generally seen as the best estimate of a property’s fair market value.

Fortunately, we had easy access to sales data, since The Times-Picayune, like many papers, publishes a weekly list of all reported real-estate transactions. Using that, a couple of computer-savvy staffers helped me create a Microsoft Excel spreadsheet of all sales for the first six months of 2003. The initial list numbered about 5,800.

The time frame for the sales was important because of what we knew about assessment practices. Typically, Louisiana assessors revalue properties after they sell, and the new number is usually close to the sales price. So it’s the properties that go unsold for decades that tend to be drastically undervalued. We wanted to look at sales that occurred after the tax assessment for that year had already been set.

Once we had the data, the next step was to clean it. We wanted only residences, so we removed all properties that we could identify as commercial, something that was frequently discernable by address or sales price. We decided to further weed out all condos or anything that looked like a multi-unit apartment complex to make the data as uniform as possible. We took out timeshares, which are common here.

Then, we took out all properties that sold for $75,000 or less, because owner-occupied properties worth less than $75,000 are exempt from taxes in Louisiana. We wound up with about 1,500 properties on the list.

Next, it was time to plug in the assessment values. Though a computer whiz might have been able to somehow match the assessments database with our file of property sales, I wasn’t able to, so this step was pretty laborious. It was more so because the city’s Website included data about previous sales in many cases, which I decided to plug in as well because I figured it might prove useful. (It did.)

After that, I started analyzing the data in various ways for instance, sorting it to see which assessments were most “off,” and to see which homeowners got the biggest tax breaks. These sorts helped clean the data further.

I eliminated houses that had clearly been bought, renovated and “flipped,” figuring these would skew the data and make the assessors look worse. For instance, if a house was bought in 2002 for $15,000 and then sold in 2003 for $150,000, I figured that was due to a renovation and it would be unfair to suggest the ho’use was wildly underassessed at $15,000.

The sorts helped pick up a couple of other oddities for instance, a house whose sales price was listed at $1.8 million, which upon further inspection I realized should have been $180,000.

Once the data was clean, it was pretty easy to analyze. The main things we tried to establish were: What is the average level of under-assessment, by district? And how much tax money is being left on the table as a result?

There are a couple of ways to look at under-assessment. One is to compare the sale price to the assessed value; the other is to compare the assessed value to the sale price. While the first figure is probably more intuitive, we wound up using the second.

In large part, that was because a leading real-estate analyst suggested that we should control for price appreciation from the past few years from our numbers. He said this after I showed him my initial analysis, which showed that homes in New Orleans were under-assessed by an average of 41 percent.

He said that since the last quadrennial assessment was due in 1999, and because the average home had appreciated in value by 31 percent since then, I didn’t appear to have much of a story.

When I thought about his reasoning, I realized he was using a different number for his denominator than I was. Perhaps an example will illustrate the point. A typical house in my survey was worth $170,000 but assessed for $100,000  which is 41 percent less than $170,000. Viewed in reverse, if that house had been worth only $100,000 in 1999, it would have had to “appreciate by 70 percent to reach $170,000. But if it had appreciated at the average rate, it would be worth only $131,000.

I decided the more sensible approach was to talk about how much buyers paid above assessed value, and then note as a caveat that part of the difference owed to appreciation since the last assessment. However, it turned out that not much of the under-assessment we found was attributable to appreciation.

When the new quadrennial reassessment was released in late 2003, it showed that the assessed value of homestead properties had only gone up 10 percent since 1999, while sales figures suggested they should have gone up by three times that much.

The analysis showed that the average person who bought a home in 2003 in New Orleans paid 70 percent over assessed value. It also allowed us to compute “error rates”for each of the seven districts in essence, showing how homeowners in some parts of the city subsidized the tax bills of homeowners in other parts. (We had to add sales from the latter half of 2003 for two districts because the initial sample sizes were deemed too small to draw such conclusions. Our final survey included at least 100 properties from each district.)

The trickiest math involved our estimates of how much money was being lost from the under-assessment. It was easy enough to figure out how much tax was lost because of the under-appraisal of the specific houses in our survey, which comprised about 2 percent of the owner-occupied homes in the city. And it was easy enough to extrapolate that number citywide.

But deciding how to treat the roughly 50 percent of homes in the city that fall under the $75,000 homestead exemption threshold was more dicey. Such homes were left out from our survey, but our data suggested that many, if not most, of the homes should not be fully exempt. (The average price for a New Orleans home sold in 2003 was $202,000. That more than half of homes are assessed for less that $75,000 suggests a problem.)

In the end, though, we opted for a conservative approach, figuring that none of the homes below the threshold should be paying taxes, even though it’s clear that many should be. We ran our methods Russ Robins, a dean at Tulane’s A.B. Freeman School of Business who said the approach seemed very safe to him. it wasn’t perfect it was overly conservative but we wanted to be careful not to overreach.

Apart from the overall trends, the data gave me perhaps some of the best anecdotes forthe story. We Used some of the more out-of-whack assessments including a French Quarter town house that had been assessed for $260,000 until it sold in 2003 for $3.5 million for the lede.

Another sort of the data led the short end of the stick. While most homeowners had properties that were under-assessed, he had purchased a home that had sold the year before and was thus assessed at about its market value of $1.2 million. He said he didn’t mind his $20,000 annual tax bill, but he resented the fact that most of his neighbors who owned comparable houses but had lived in them for a long time were paying far less.

Said the homeowner, “It’s totally unfair. If I can afford to buy this house, I can afford the taxes other people who have houses like this.”

Contact Gordon Russell by e-mail at [email protected]