All too often Corvallis operates as if everybody in Corvallis had a great deal of money. Or to put it another way, all too often the powers-that-be think money grows on trees. Therefore, I thought it would be interesting to see just how much money Corvallis citizens actually earn. I used data on earnings by occupation from the U.S. Bureau of Labor Statistics (BLS), income tax data from the Oregon Department of Revenue (DOR), and the 2008 American Community Survey (ACS) 1-Year Estimates for the Corvallis, OR, Metropolitan Statistical Area (MSA) to estimate how well Corvallis residents are doing. I would like to have had data like that contained in the DOR data for every taxpayer in Corvallis; but, I could not find such a data set.
Some people in Corvallis do very well. Most do not do all that well. And, too many people do not do well at all. I estimate that the median income is between $50K and $60K per household, $30K per individual, and that about 30% of the population of Corvallis can be considered poor and eligible for food stamps and other poverty programs. Despite the great recession, the upper middle is probably still doing OK and top quintile is probably still doing very well -- if they still have jobs. When the powers-that-be want to impose more taxes on the citizenry, they need to know that additional taxes will be a hardship for the poorer people and discourage the richer people from working harder. Imposing more taxes might encourage the top earners to leave Corvallis.
According to data collected by the BLS for the Occupational Employment Statistics (OES) May 2008 Survey, the median annual wage in the Corvallis MSA -- which is actually Benton County -- is $30,620. In other words, one-half of the people earn less than $30,620 -- not a great deal of money these days. Ninety percent of the people earn less than $74,230 -- considerably better than the median but still not a great deal of money. These data do not show the bigger picture because they are wages for all occupations and do not include money from pensions or investments. Nor do the data contain the tails of the wage distribution. We have highly paid professionals skewing the distribution which is evident by the mean annual range being $40,030 -- to the right of the median. The OES data can be seen here. The descriptions of the fields in the data can be seen here.
In an attempt to look at the tails of the wage distribution, I estimated the total distribution of annual wages using the means of the annual wages and standard errors of the means of individual occupations.
I estimated the distributions of each mean for of each occupation and summed the distributions weighting each distributions by the number of people employed in each occupation. As can be seen by perusing the OES data, all the means and standard errors by occupation are not available. A graph of the estimated cumulative distribution function is shown here along with a plot of the percentiles of the original data -- which is not available -- for all occupations in Corvallis.
As can be seen, the estimated distribution of the weighted means is to the right of the plotted percentiles. The left tail is not noticeably left of an annual wage of about $17K. However the right tail is noticeable up to about $220K.
The Oregon DOR publishes various tables of Oregon income tax data by sources of income, counties and selected cities.
The latest data are for 2007. The data seem to be organized in such a fashion that individual people cannot be identified. For example, in a very small city, individuals with very high incomes could possibly be identified. I used the data for Benton County which can be found here. I estimated the mean adjusted gross income (AGI) for each filer by quintile with the fifth quintile divided into three parts. A graph of those means by quintile is shown here.
As is rather obvious, the tails of the two distributions -- the BLS OES Data and the DOR - Tax Data -- are very different. 
The top 1% of Benton County, as per the DOR - Tax Data, is doing extremely well relative to everyone else. And, the next 4% is extremely comfortable.The bottom quintile is not doing very well. The mean AGI for the bottom quintile is about $2,600. I find it difficult to believe that people with AGIs that low can survive on their AGIs. Presumably these people are on public assistance, receive tax credits of some sort, or can be claimed as dependents on someone else's tax return. The graph shown here displays the average number of exemptions the filers in each quintile claim. As can be seen, the average number of exemptions of the first quintile is less than one indicating that a sizeable number of filers are claimed on someone else's return.
The graph shown here displays the lowest four quintiles allowing the reader to see the relative picture of the not-so-rich more easily.
The U.S. Census Bureau publishes data profiling MSAs.
One of the data sets is income with in the past twelve months. A graph showing the cumulative distribution function of that income for households, families, and nonfamily households is shown here.The definitions of family and family-households, taken from the Census Bureau web site, are in notes 1 and 2 below. For more information, the reader should visit the Census Bureau web site. The graphs are plotted up to $200K. The upper limit on income is not presented in the data. As can be seen, nonfamily households do more poorly than families.
The income data within the past twelve months is presented as a set of means and standard errors of households that earn the amount of income in the various bins (e.g. $0-$15K, $15K-$25K)..
The graph of cumulative distribution functions was plotted using the means of that data; the errors are not reflected in the plots. The graph shown here is a simulation of the cumulative distribution function for the income for households. The dark lines are the 90% confidence intervals on the cumulative distribution function. I created the covariance matrix of the number of households in each bin assuming that the correlation coefficients between any two bins were equal.
There seems to be a correlation between income and exemptions in the DOR data.There also seems to be a correlation between household size and income in the ACS data as shown here. The smaller nonfamily households seem to have less income than families.
One of the ACS data sets is the ratio of income to the poverty level in the past twelve months.
The universe is the population for whom the poverty status is determined -- 75,392 people in the Corvallis MSA. The data is presented as set of means and standard errors of the number of people whose ratio of income to the poverty level falls in various bins (e.g. under 0.50, 0.50 to 0.74, 0.75 to 0.99). The graph shown here is a simulation of the cumulative distribution function for the ratio of income to poverty level for the population. The dark lines are the 90% confidence intervals on the cumulative distribution function.The graph is plotted up to a ratio of five. The upper limit on the ratio is not presented in the data. I created the covariance matrix of the number of people in each bin assuming that the correlation coefficients between any two bins were equal. As can be seen, about 18% of Corvallis lives below the poverty level and about 33% live below two times the poverty level.
The amount of money a household had to earn in 2008 to be in poverty is shown here.
Federal programs using the guidelines (or percentage multiples of the guidelines — for instance, 125 percent or 185 percent of the guidelines) in determining eligibility include Head Start, the Food Stamp Program, the National School Lunch Program, the Low-Income Home Energy Assistance Program, and the Children’s Health Insurance Program.
One of the ACS profiles published by the Census Bureau is the number of households on public assistance or food stamps. According to the this profile, the ratio of households receiving public assistance or food stamps is 7.91% with a standard error of 1.49%. The 90% confidence interval is 5.47% to 10.36%.
A measure of poverty is a family's eligibility for food stamps.
The maximum gross and net incomes required to receive food stamps are shown here. I refer the reader to the Department of Agriculture web site for the difference between gross and net incomes. The maximum incomes required to be eligible for food stamps seems to be at variance with the ratio of households receiving public assistance or food stamps making me wonder how many people don't take food stamps if they are eligible, or won't admit to being on public assistance or food stamps. There are probably quite a few one and two person households that are eligible for food stamps. People in the first two quintiles of income -- other than dependents -- are probably eligible for food stamps. And, there are probably people in the middle quintile who are also eligible for food stamps.
The mean (AGI) of the middle -- third -- quintile is $35,236; a figure between the mean and median of the BLS OES data. However, these data sets are very different from each other. The BLS OES data is that of individuals in occupations. The DOR Tax Data is that of income tax filers -- single, joint, etc. And, the ACS data sets are different from the other two data sets. The ACS data when compared to the other two data sets leads me to believe that the lower income people are the nonfamily households.
The data sets are collected in different, but consecutive, years. However, they are close enough in time for me to believe that the effects of inflation are not important relative to the other differences in the data sets.
The grand AGI mean of the DOR Tax Data is $59,801 and the mean of the BLS OES data is $40,030 -- which, ignoring the number of exemptions and filing status, is presumably AGI minus annual wages, $19.8K. The average difference between AGI and wages in the DOR Tax Data is $22.8K -- a figure reasonably close to $19.8K. The mean AGI for the city of Corvallis is $58,566 -- slightly less than the Benton County AGI grand mean telling me that the mass of the Corvallis distribution of AGI is a little to the left of mass of the Benton County distribution. However, the average number of exemptions and the very low AGI for the first quintile make me wonder how many of those filers are young people being primarily supported by their parents. If one drops the first quintile from the DOR Tax Data, the mean AGI is $74,101 -- a $14K improvement.
The data sets are for the years 2007 and 2008. 2007 is just before the great recession started. We keep hearing of companies in Corvallis that have been closing or down-sizing and of people in Corvallis who have lost their jobs since the great recession started. Therefore the economic situation may be worse than analysis of these data sets indicate.
As I said earlier, it would have been very convenient to have had data like that contained in the DOR data for every taxpayer in Corvallis. But I don't, so I have to go with what I have. I can't make a best estimate -- unbiased with minimum variance -- with a single coherent data set. All I can do is make informed estimates -- informed by analyses of these data sets. I estimate that the median income is between $50K and $60K per household, $30K per individual, and that about 30% of the population of Corvallis can be considered poor -- eligible for food stamps and other poverty programs. The upper middle is probably still doing OK and top quintile is probably still doing very well -- if they still have jobs.