The taxman knows

Let me tell you how it will beThere’s one for you nineteen for me‘Cos I’m the taxman Yeah I’m the taxmanDon’t ask me what I want it for (ah ah, Mr. Wilson)If you don’t want to pay some more (ah ah, Mr. Heath)--The Beatles

Tax departments and knowledge management--now there’s a combination sure to drive terror into the hearts of most of us who view the "taxman" in Beatle-esque terms. Whether we like it or not, the taxman is getting smarter, and there is a serious, concerted effort in most state revenue departments to manage the knowledge derived from processing the personal and corporate returns that inundate them annually.

To get an idea of the relative dimensions of that tidal wave of paper, the office of the Texas Comptroller receives as many as 144,000 returns a day during the tax monsoon, which can equal 360,000 pages. Over the course of a year, that equates to more than 13 million documents. And those are only corporate returns.

A typical pre-KM scenario for state revenue organizations starts with a mass infusion of temporary workers at the beginning of tax season to attempt to stem the tide. In some cases, there still are insufficient resources to capture all the information on the forms. Why? Three factors come into play--the increasing complexity of tax law that requires an endless number of worksheets, forms, cross calculations, table lookups, conditional calculations and more; an increasing population; and decreasing state budgets that push departments to do more with less.

Armed with incomplete information, auditors have a difficult time doing their work. If they suspect something is amiss, they must go to either the original paper or a microfilm copy of the return to get the information they need. That leads to many missed opportunities to collect additional revenues and much wasted time.

Furthermore, other departments in state government would love to get their hands on the information in tax returns. For example, economic development offices could glean useful information about how well corporations are doing or they could identify disturbing trends that could benefit from preventative action. None of that is possible if we capture the bare minimum of information.

Now let’s look at the possibilities with knowledge management. If we could capture all the information and build adequate databases, we could use business intelligence tools, such as CART (Classification and Regression Trees) or CHAID (Chi-squared Automated Interaction Detection) trees, which can build profiles of returns that are likely fraudulent. That is being done in insurance companies to learn, for example, that if X and Y are true on a particular claim, there is a Z% chance that the claim is fraudulent. The result over time is less fraud and more revenue, and fewer people who are audited unnecessarily. The economic development office could also use data mining, online analytical processing (OLAP) and other business intelligence tools to turn information into knowledge and make it actionable.

Missouri wants to tie tax and census data with GIS systems to build knowledge systems that can identify, for instance, the major source of tax revenue by locale across the state. Income distribution patterns, age distribution patterns, all of those necessary pieces of planning information would be available at a glance. Child service agencies, Secretary of State offices, the list goes on. The one common factor that makes any of that possible is a solid database of information, culled from the tax returns.

That’s why data capture tools like forms processing are so important. Not only do they solve the immediate problem of doing more with less, they also unlock possibilities downstream in terms of sharing and using knowledge locked in the returns.

The next step is to empower the users and extend knowledge gathering to the taxpayers themselves. Michigan saved more than $300,000 last year by providing electronic filing, and response was up 45% over last year. In the corporate community, many organizations use computer applications to build their filings. What do they do next? They dutifully print the required number of copies and send them to the taxman, so that the data--already entered once--can be re-entered by hordes of temporary data entry clerks. How much sense does it make to install forms processing in situations where a large number of forms submitted are electronically created? Several years ago, a federal agency was preparing just such a system until it found out, at the behest of a consultant, that 60% to 80% of its forms were generated electronically. The system requirements quickly morphed to include electronic filings and a host of incentives to get corporations to use them.

Core technologies can make the taxman more efficient and effective, and underscore why the taxman not only wants your money, but also your information.