THE ECONOMICS OF BANK ROBBERIES IN NEW ENGLAND 1. Kimberly A. Leonard, Diane L. Marley & Charlotte A. Senno

Transcription

1 THE ECONOMICS OF BANK ROBBERIES IN NEW ENGLAND 1 The Economics of Bank Robberies in New England Kimberly A. Leonard, Diane L. Marley & Charlotte A. Senno The University of Rhode Island, STA308 Comment [N1]: Very good work. Points: 75 / 75

2 THE ECONOMICS OF BANK ROBBERIES IN NEW ENGLAND 2 Abstract The topic of the paper was to prove an increase in the number of bank robberies in the six New England states of Maine, New Hampshire, Vermont, Massachusetts, Connecticut and Rhode Island as years increased since Seven years worth of data was collected on these six states in the areas of bank robberies, unemployment, bankruptcy and state population. Paired T-tests for the following pairs: , , , , , were calculated to determine if there was an increase in bank robberies and unemployment. The team found no significant increase in the number of bank robberies over the seven years; however, we found an increase in unemployment in years Using a chi-square test for independence, we found that when looking at bank robberies and bankruptcies the variables years and states are not independent and unemployment rates were independent. We found no relationship between unemployment rates and bank robberies using a linear regression. Therefore, our hypothesis of bank robberies increasing over time has been disproved. Introduction Recent Rhode Island media coverage has included a large number of stories on local bank robberies. Media focus also remains largely on the current economic condition. But are the two issues connected? Our hypothesis is that the frequency of bank robberies increases as economic

3 THE ECONOMICS OF BANK ROBBERIES IN NEW ENGLAND 3 stability decreases. We chose to define economic stability as a combination of two variables: unemployment rate and bankruptcy filings. An increase in either variable is taken to indicate a weaker economy, whereas lower numbers are representative of more stability. In order to prove or disprove our hypothesis, the research team has chosen to collect data on each of the six New England States: Connecticut, New Hampshire, Massachusetts, Maine, Rhode Island and Vermont. For each state, the number of bank robberies, percentage of unemployment, and number of bankruptcy filings has been collected for a seven year period, from (We planned on a ten year period, but FBI statistics were unavailable from However, this actually helps to minimize error, as the economic condition around 9/11 would certainly be an outlier.) The state population by year is also included for use in comparison and analysis of data. Lit Review Comment [N2]: Don t use abbreviations Unemployment figures are drawn from the US Department of Labor (USDL) which provides seasonally adjusted new filings and persons currently receiving benefits. The percentages given do not include the number of persons who have exhausted their benefits, under employed individuals, those who are no longer looking for work nor the economic class of claimants. According to the USDL Seasonal adjustment is a statistical method for removing the seasonal component of a time series used when analyzing non-seasonal trends. Whereas, not seasonally adjusted reflects the actual current data. It is normal to report not seasonally adjusted data for current unemployment rates. Seasonally adjusted data may be used for the longer term comparison.

4 THE ECONOMICS OF BANK ROBBERIES IN NEW ENGLAND 4 Bankruptcy is the The legal process involved: the administration of an insolvent debtor s property by the court for the benefit of the debtor s creditors. Filing by a debtor is called voluntary bankruptcy; involuntary bankruptcy is declared by the court upon petition by creditor (Britannica Concise Encyclopedia). We have chosen to represent bankruptcy as a total number of filings per state, disregarding the category the bankruptcies were filed in. Our bankruptcy data was collected from uscourts.gov. The FBI Bank Crime Statistics electronic database lists the number of bank robberies, bank burglaries, bank larcenies and bank extortions per year since calendar year Definitions obtained from the US Attorneys Manual define these illegal activities as: A bank robbery is the taking or attempted taking by force, intimidation, or extortion, of any property in the care, custody, control, management or possession of any bank, credit union, or savings and loan association. Typically, some form of violence is demonstrated. In contrast, a bank larceny is the taking of property in nonviolent face-to-face encounter and bank extortion arises where, by telephone call or other communication, an extortionist conveys a threat to a bank official. Where robberies take place during an institutions working hours, burglaries include breaking and entering when the institution in question is closed. For our data and analysis, we defined bank robberies as an inclusive term: one value representing a combination of bank robberies, larcenies, extortions and burglaries. Methods The team collected yearly data from for the six New England states. Data was collected in the following areas: number of bank robberies, unemployment rate and bankruptcy filings. State population was also collected. This information is outlined

5 THE ECONOMICS OF BANK ROBBERIES IN NEW ENGLAND 5 below, in both tables and graphs. Bank Robberies State CT ME MA NH RI VT Unemployment Rate State CT ME MA NH RI VT Bankruptcy Filings State CT 12,246 11,423 15, ME MA NH RI VT U.S. Census State Pop State CT ME MA NH RI VT

6 Bank Robberies Population THE ECONOMICS OF BANK ROBBERIES IN NEW ENGLAND 6 New England State Populations, Year CT ME MA NH RI VT N.E. Bank Robberies, CT ME MA NH RI VT Year

7 Percent Unemployment (%) # of Filings THE ECONOMICS OF BANK ROBBERIES IN NEW ENGLAND 7 Bankruptcy Filings, ,000 20,000 10, Year CT ME MA NH RI Unemployment Rates, Year CT ME MA NH RI VT Mean, Median, Mode Summary statistics: Bank Robberies Column n Std. Dev. Std. Err. Median Range Min Max Q1 Q3 Mean Comment [N3]: When presenting your data always use the same number of decimal points in a column. Unless the numbers are very small a two or three decimal accuracy should be enough

8 THE ECONOMICS OF BANK ROBBERIES IN NEW ENGLAND 8 Summary statistics: Unemployment Rates Column n Mean Std. Dev. Std. Err. Median Range Min Max Q1 Q Summary statistics: Bankruptcy Filings Column n Mean Std. Dev. Std. Err. Median Range Min Max Q1 Q Summary statistics: State Population Column n Mean Std. Dev. Std. Err. Median Range Min Max Q1 Q

9 THE ECONOMICS OF BANK ROBBERIES IN NEW ENGLAND 9 Population is consistent for each state over the seven year period. The shape is fairly consistent with unemployment with the exception of New Hampshire in fiscal year Bankruptcy filings have a consistent shape for all states; however, Massachusetts and Connecticut have significantly higher rates. The number of bank robberies reflects a relatively similar shape with outliers in Massachusetts and Connecticut shows elevated rates. Comment [N4]: Label your graphs / charts and then refer to them when analyzing.

10 THE ECONOMICS OF BANK ROBBERIES IN NEW ENGLAND 10 The median in the bank robbery box plot is relatively the same throughout and the IQR which represents 50% of the data is also stable across the years. The outliers are from the bank robberies committed in Massachusetts as seen on the line graphs. The median in the box plot reflecting bankruptcy filings are consistent except in 2005 and Further research shows that a change in bankruptcy laws which made filing for bankruptcy more difficult drove the number of filings up in 2005 followed by a much lower amount in 2006, once the law went into effect. This also resulted in the drastic lowering of the IQR in 2006, which remained fairly low for the rest of the years we observed. Analysis & Results Paired T Test Hypothesis test results: Bank Robberies μ 1 - μ 2 : mean of the paired difference between 2003 and Tcv = Hypothesis test results: Bank Robberies μ 1 - μ 2 : mean of the paired difference between 2004 and

11 THE ECONOMICS OF BANK ROBBERIES IN NEW ENGLAND 11 Hypothesis test results: Bank Robberies μ 1 - μ 2 : mean of the paired difference between 2005 and Hypothesis test results: Bank Robberies μ 1 - μ 2 : mean of the paired difference between 2006 and Hypothesis test results: Bank Robberies μ 1 - μ 2 : mean of the paired difference between 2007 and Hypothesis test results: Bank Robberies μ 1 - μ 2 : mean of the paired difference between 2008 and The team performed a series of paired T-tests for the data collected on number of bank robberies. Each hypothesis test resulted in a. This means we cannot reject the null (that bankruptcies do not change between each 2 year period). There could be some change, but we did not have sufficient evidence to prove bank robberies increased over time.

12 THE ECONOMICS OF BANK ROBBERIES IN NEW ENGLAND 12 Hypothesis test results: Unemployment Rates μ 1 - μ 2 : mean of the paired difference between 2003 and 2004 Tcv = Hypothesis test results: Unemployment Rates μ 1 - μ 2 : mean of the paired difference between 2004 and Hypothesis test results: Unemployment Rates μ 1 - μ 2 : mean of the paired difference between 2005 and Hypothesis test results: Unemployment Rates μ 1 - μ 2 : mean of the paired difference between 2006 and

13 THE ECONOMICS OF BANK ROBBERIES IN NEW ENGLAND 13 Hypothesis test results: Unemployment Rates μ 1 - μ 2 : mean of the paired difference between 2007 and REJECT Hypothesis test results: Unemployment Rates μ 1 - μ 2 : mean of the paired difference between 2008 and < REJECT The team performed a series of paired T-tests for the data collected on unemployment rates, as well. The hypothesis test could not be rejected for intervals from 2003 to This means there was not sufficient evidence to prove that unemployment rates changed over this time. From and , hypothesis tests resulted in rejecting the hypothesis. This means unemployment rates probably did change during these 2 year periods, but we cannot determine if they increased or decreased, as we performed a one sided test. This is most likely because of NH, as you can see in the unemployment graph; NH s unemployment varies from the normal distribution from 2007 to Contingency table results: Bank Robberies Statistic DF Value P-value Chi-square < CRITICAL VALUE: REJECT Comment [N5]: A one-sided test would tell you if it increased or decreased. In your case since you are assuming U1 - U2 < 0 it implies that U1 is less than U2. A rejection of the null implies that U1 is less than U2. However, I doubt if the program you used to do the test actually did a one sided test. So unless you compared to the CV your self it was a twosided test and your alternative hypothesis should have been stated as a two-sided test U1 U2 <> 0

14 THE ECONOMICS OF BANK ROBBERIES IN NEW ENGLAND 14 Contingency table results: Unemployment Rates Statistic DF Value P-value Chi-square Contingency table results: Bankruptcy Filings Statistic DF Value P-value Chi-square < REJECT Chi-square analyses for bank robberies, unemployment and bankruptcy filings, were conducted to determine if these values were independent of year and state. For bank robberies and bankruptcy, the chi-square test result was to reject the null. This means these values are not independent of year and state. In contrast, unemployment rates were independent of year and state. Simple linear regression results: Dependent Variable: Bank Robberies Independent Variable: Unemployment % Bank Robberies = Unemployment % Sample size: 42 R (correlation coefficient) = R-sq = Estimate of error standard deviation: Parameter estimates: Parameter Estimate Std. Err. Alternative DF T-Stat P-Value Intercept Slope Analysis of variance table for regression model: Source DF SS MS F-stat P-value Model Error Total

15 THE ECONOMICS OF BANK ROBBERIES IN NEW ENGLAND 15 H 0 : B = 0 H a : B 0 The linear regression test resulted in a P value of therefore the null hypothesis is not rejected. Our findings reflect that there is no relationship between unemployment and bank robberies. Discussion/Conclusion The point of this research was to determine if there is a relationship between the number of bank robberies and the economic factors of unemployment rates and bankruptcy filings. Our hypothesis was that increased unemployment and bankruptcy would result in increased prevalence of bank robberies. We believe the recent economy has declined and the number of bank robberies reported in the media has increased. The paired-t test disproved an upward trend in bank robberies. There was no significance in unemployment, either, until 2007, where it remained significant in 2 year periods until This is consistent with our observation of increased unemployment in recent years.

16 THE ECONOMICS OF BANK ROBBERIES IN NEW ENGLAND 16 Our linear regression test tested for correlation between unemployment rates and number of bank robberies. After performing the test, we determined there is no correlation between these two variables based on the evidence presented. Therefore, all of our testing has proven our initial hypothesis was incorrect---there is no notable relationship between bank robberies and unemployment/bankruptcy based on the data we collected and analyzed.

17 THE ECONOMICS OF BANK ROBBERIES IN NEW ENGLAND 17 References Agee Sara, (2006)- M&M Math Science Buddies retrieved December 13, Federal Bureau of Investigation Bank Crime Reports. ( ) Retrieved October , New England Unemployment Statistics. ( ) Retrieved October 20, Rhode Island State Courts, Bankruptcy Statistics. ( ) Retrieved November 2, 2010 from Schiffman, Barry. (1985, April 26). Unfruitful crime. Wall Street Journal (Eastern Edition), p. 1. Retrieved December 13, 2010, from ABI/INFORM Global. (Document ID: ). State of Maine Bankruptcy Statistics. ( ). Retrieved November 12, 2010 from United States Bureau of Economic Development. ( ). Retrieved October 2, 2010 from United States Local Area Unemployment Statistics. (LAUS). ( ). Retrieved November 20, 2010 from United States Economic Summaries by States. ( ). Retrieved November 20, 2010 from United States Department of Labor Statistics. New England Information Office. ( ) Release Date September 28, 2010 Retrieved October 20, 2010 from United States, State Population. ( ) Retrieved October , from

18 THE ECONOMICS OF BANK ROBBERIES IN NEW ENGLAND 18 Zara, Georgia; Farrington, David P.. Criminal Behaviour & Mental Health, Oct2010, Vol. 20 Issue 4, p , 17p, 6 Charts, 2 Graphs; Retrieved December 13, 2010 DOI: /cbm.763