COR OVERTIME COMPENSATION: BENFORD LAW FRAUD DETECTION HIGHLY PROBABLE

In response to the following insider tip alleging fraud in overtime compensation in the fire department, with corresponding but less detailed information provided involving alleged similar abuse in the parks and police departments, and a recommendation from an individual with extensive law enforcement and white collar criminal investigation experience, further forensic accounting and fraud detection analysis was performed on the city’s 2008 and 2009 overtime compensation information provided in an Open Records Request in conjunction with the original blog on this matter entitled “Over-the-Top Overtime Compensation – Is There a Draft in Here” as detailed below:

— Forwarded Message —
From: COR Employee currentemployee@cor.gov
To: editor@cornews.net
Sent: Friday, August 21, 2009 2:59:39 PM
Subject: cornews: Car Allowances
This is an enquiry e-mail via http://www.cornews,net/from: COR Employee currentemployee@cor.gov

If you opted to make an open recrods request for the car allowance rules and mileage rules in the COR human resources policies and procedures you would see what King Keffler was doing is against his own rules (ie. using a city vehicle if you get a car allowance).The only time you can do this is if your car is broke or not large enough for the job. In addition, the on duty shift commanders for the fire department have a personal driver, a city Fire Department decaled SUV, and all three get a $492 car allowance. The Fire Chief often takes the fire department cars (Admin 1 or 2) for his own use while he get a personal car allowance as well. This city is so corrupt that I am disgusted to work here. If you really want to see corruption, request the overtime records of the fire department. Over 1 million in overtime for the FD alone.

Benford’s Law Principle:
http://en.wikipedia.org/wiki/Benfords_law

Benford’s Law, also called the first-digit law, states that in lists of numbers from many (but not all) real-life sources of data, the leading digit is distributed in a specific, non-uniform way. According to this law, the first digit of 1 occurs almost one third of the time, and larger digits occur as the leading digit with lower and lower frequency, to the point where 9 as a first digit occurs less than one time in twenty.

As shown on the chart below depicting the Benford’s Law probability distribution, along with the derived mathematical formula, with the y-axis representing the probability of occurrence from the smallest (low) to largest (high) and the x-axis representing single digit or combinations of numbers from smallest (left) to highest (right), the Benford’s Law distribution of numbers is a gradually sloping downward curve that allows one to reasonably predict based upon statistical probabilities how often such randomly occurring single digit, or combinations of digits, would normally occur in a valid sample of data:

In essence, the Benford’s Law principle of statistical probabilities is based upon the inherent random frequency/occurrence of numbers and combinations thereof. The lower the number, the higher the probability of that number occurring in a statistically valid sample of data. Lower numbers such as one (1) or two (2) for first or second digits will occur more often on a random basis than the numbers seven (7) or eight (8). Similarly, combinations of lower numbers, such as the combination of 12 consisting of the first digit of 1 and the second digit of 2 will occur more frequently than combinations of higher numbers, such as the combination of six (6) and seven (7) or combination of 67.

Benford’s Law has been found to apply to a wide variety of data sets, including electricity bills, street addresses, stock prices, population numbers, death rates, lengths of rivers, physical and mathematical constants, processes described by power laws, voting/election fraud, and most importantly, forensic accounting and economics analysis, fraud detection in predictive analysis, and accounting and expenses fraud. Certified fraud examiners use this program or a derivative to detect government or business/corporate fraud. In the U.S., evidence based on Benford's law is legally admissible in criminal cases at the federal, state, and local levels.

Harvard Business School Journal of Accountancy (HBSJA) and AICPA – Fraud Buster Detection Program:

http://www.journalofaccountancy.com/Issues/2003/Aug/TurnExcelIntoAFinancialSleuth.htm

The HBSJA in conjunction with the Association of Independent Certified Public Accountants (AICPA) have put together a simple to use and understand fraud detection Excel spreadsheet entitled “FRAUD BUSTER” based upon Benford’s Law that will allow one to perform forensic accounting and fraud analysis of financial data to determine whether or not there is a high probability of fraud being committed. After entering the data and running the program, if single first or second digit numbers or combinations of first and second digit numbers fall well outside the expected norms or frequency of occurrence, either under (less than expected) or over (more than expected), there is an increased probability that fraud is being committed. The higher the deviation, above or below, the higher the probability of fraud. The Fraud Buster program will detect either outright fraud or attempts to cover up fraud, as both acts will fall outside the normal realm of random occurrence/probability.

Benford’s Law/Fraud Buster Forensic Analysis Results of COR Overtime 2008-2009 Compensation:

Although Benford’s Law has certain limitations and is not effective in detecting fraud for all financial data, such as instances where the data set is too small, where data include built-in minimums and maximum’s, or where data sets included assigned numbers, none of these situations appeared to be applicable involving the COR overtime compensation and hours claimed information provided via an Open Records Request. This information consisted of 2,474 pieces of data consisting of overtime salary paid and hours claimed constituting 2/3 of the total overtime paid by the city over the past 2 years. Furthermore, this data representing overtime paid to 409 city employees in the parks, fire, and police departments represents nearly 40% of the total city current employee base. Thus, it appears that the analysis presented below is based upon a statistically valid sample pending verification and further review by law enforcement officials and statistical experts.

After compiling, sorting, and then segregating the data by department and running various Fraud Buster test scenarios based upon overtime hours claimed, overtime salary paid, and combinations thereof for one or both years to identify and potentially isolate possible areas of abuse, preliminary but fairly conclusive analysis using the Benford’s Law Excel Fraud Buster detection program on the OT data, as detailed below on the various Fraud Buster program graphs depicted for First Digit, Second Digit, and First Two Digit tests in comparison to HBSJA example/sample fraud case graphs, indicates that there is a high probability of significant and widespread fraud being committed. Similar to the HBSJA sample fraud case graphs, statistical deviations above or below the expected norm of 3-5X, as well as highly unusual recurring and nearly identical/level patterns of outlier first and second digit data sets from the lowest to highest pairs, were noted. Further analysis, sampling, and more specific testing scenarios could possibly be run to help further isolate the suspected areas of abuse.

Along with the questionable “plug” figures found in the fire, parks, and police overtime records (matching cents figures for both the gross compensation received and overtime salary paid for certain individuals for two consecutive years) that by themselves violate the Benford’s Law principle, are highly suspicious, and worthy of further investigation, these Benford’s Law Fraud Buster tests further support the anonymous tips provided from insiders in the fire, parks and police departments that fraud and/or theft is being committed. This is further supported by the high amounts of overtime compensation being provided from one year to the next, excessive hours being claimed, and the high levels of recurring overtime being paid from one year to the next as a percentage of total compensation, all of which raise further question.

Summary:

Given that the city council and city manager’s office has been advised of this matter and refuses to launch an investigation into the alleged overtime abuse/fraud and theft, this information is being provided to law enforcement officials at various levels whom should conduct a full and complete audit of all overtime records of these three departments over the past 3-5 years. Whistleblowers and witnesses are encouraged to come forward with additional information, and they may be eligible for a sizeable cash reward being offered. More importantly, it is the right, honest, just, and honorable thing to do.