A true story about using XML forms and DB2 to process millions of tax returns
IBM’er Brent Cooley talks about how NY State used XML forms and DB2 to process millions of tax returns and reaped significant savings and performance boosts. Enter Brent:
How do I cut form processing costs, speed application development, and at the same time, detect mistakes and questionable claims more effectively?
Choking on Paper
From insurance to real estate, health care to government, many organizations are mired in an endless wave of paper. Envelopes must be opened and paper forms scanned (or far worse, keyed by hand) before any amount of electronic processing can occur. The deluge of paper forms is seen by most as a drag on already-tight budgets and stressed staffing levels as well as being one of the biggest roadblocks to controlling cost.
Due to the promise of cost savings, processing efficiencies, and enhanced capabilities such as predictive analytics, organizations far and wide are looking to get out of the paper rut and onto the e-form highway.
The Tax Man Cometh
One such organization, the New York State Department of Taxation and Finance (NYS DTF), has been undergoing a massive modernization effort. This organization processes over 10 million personal income tax returns annually, most of which arrive between mid-January and mid-April. Shortly after filing tax returns, taxpayers expect that any refund due is paid in a timely manner. With budgets tightening and staffing levels being reduced, electronic form processing was the answer. Since a very high percentage of taxpayers already fill out their tax forms electronically, it made sense to process them completely electronically as well.
Enter XML and DB2
One of the major considerations for any large e-forms effort is the level of maintenance required to support new forms, new data fields, and changes in business process. Since we all know tax filing can be somewhat complicated with many forms, schedules, and documentation changing from year to year, an e-forms solution to support tax processing needs to be flexible. For NYS DTF, XML was seen as the next frontier with its structure able to be modified from year to year with minimal impact on existing applications. Additionally, its ability to be paired with layout logic (transformations) to construct e-forms on-screen which look identical to a paper counterpart was an easy sell to business users.
Since NYS DTF was already a relational DB2 user, the premise of native XML processing within the DB2 database engine seemed like a no-brainer. Over the course of a multi-year effort, NYS DTF built a system which accepts tax returns electronically, converts, processes, and stores them as XML documents. From there, business exceptions can be handled via workflow, taxpayer accounts adjusted, questionable tax returns sniffed out, and refunds issued. The ease by which application developers can fetch and retrieve the native XML from DB2 should not be taken lightly. With XPath and XQuery support, fetching a few values from a large document is quick and easy, avoiding the overhead of parsing the entire document within the application. The functions available in the DB2 pureXML engine allow for tremendous flexibility, even with data stored in relational tables. With pureXML functions such as XMLFOREST and XMLAGG, queries can be written to turn even the scariest of relational structures into well-formed, easily processed XML content.
XML forms and DB2 helped to provide a winning solution for NYS DTF and the taxpayers and constituents of New York State. Since implementing XML-centric tax return processing, over 50 million XML documents have been processed and stored and $20 million dollars in paper handling, printing, and personnel costs have been saved. A harder to measure amount of savings in application development and maintenance may be similarly impressive.
If you are a Gartner client, click here to read a case study on how New York Saved Millions by Using E-Forms to Process Complex Tax Filings (make sure you’re logged into Gartner first).
If you are not a Gartner client or you are an IBM employee, please contact SmarterQuestions on Twitter @SmarterQuestion for further information on this topic.