Members of the Economic Research Institute of Erie have expertise in:

• forecasting of business and economic variables
• the Erie economy
• regional economics
• economic development
• labor economics
• international economics
• the Chinese economy

Most especially, we have experience in finding and analyzing data. That’s what we do!

Some of the studies that ERIE has undertaken include:

• estimates of productivity of Erie’s workers through time and across industries, compared to the nation.
• estimates of brain drain and brain gain for Penn State graduates, from Erie County.
• creation of a sophisticated model to forecast total Erie employment, as well as employment in durables and nondurables manufacturing, total manufacturing, and non-manufacturing industries.
• creation of a forecasting model to help manage inventories for a local tire retailer, cutting inventory costs.
• three studies of philanthropic giving in the Erie area, and one in the Berks County (Reading) area, sponsored by leading non-profit agencies.
• a sophisticated econometric model to estimate the cost of living in all 67 counties of Pennsylvania for a state government agency.
• an examination of the effect of Erie’s changing industrial structure on the severity of its business cycle.

We can also do presentations in the community about our work. Typical presentations might include an overview of the Erie or national economies, a discussion of productivity in Erie compared to other places, a discussion of the Chinese economy, or local philanthropy. While our schedules are often pretty full, we may be able to accommodate your needs for a speaker. Give us a call!

While we specialize in economics, we have colleagues at the Black School of Business at Penn State Behrend who have expertise in the other areas of business, such as marketing, MIS, management, finance, accounting, business strategy and policy, and statistical analysis. We may be able to help you find the skills you need.

We also have access to some excellent students, if you are looking for a good (paid) intern to help do a project that you just don’t have time to do yourself. This is an especially good way to check out a potential employee on the job, if you’re thinking about hiring in the future. Our students are trained in data handling and analysis, and participate in several hands-on projects with real-world data in our program, and they’re used to working with many of the standard computer software packages. We have former students in major corporations and government agencies around the country, and their employers tell us they are well-prepared for the world of work.

Contact us at 814-898-7150 or klouie@psu.edu. Maybe we can help! Give us an overview of your needs and we can work up a proposal for you.

Downloading data on the ERIE site is easy and fast.

Step One: Highlight the data below the chart by holding down the left mouse button and dragging. Once the data is highlighted you may let go of the mouse but you cannot click anything until you have the information copied.

Step Two: Hold the Ctrl key down then press the letter C. (This will COPY the information onto the CLIPBOARD.)

Step Three: Open MS Excel or whatever spreadsheet software you prefer.

Step Four: Right click on an EMPTY cell

Step Five: Select Paste from the menu that pops up near the cursor.

Grabbing graphs for use in MS Word or any other compatible software is fast and easy.

Step One: Hold the Alt Key Down and then press the PAUSE / BREAK key located in the upper right part of your keyboard.

Step Two: Go to the document that you are going to use to paste it into and hold shift then press the INSERT located in the upper right part of your keyboard.

For each data series our website presents the source of the data, typically with a direct link when possible. That lets you go straight to the horse’s mouth, if you’d like. In many cases these sites will present more detailed information about how they collect and process their data, and you can decide for yourself if they are trustable. Many (most?) of the data on the ERIE site come originally from government statistical agencies, which typically have a good reputation for presenting accurate data that are not subject to misrepresentation for political purposes. This is not to say that they never make an error, but that their history has typically been quite positive. And while the data are not perfect, they are typically among the best available for measure economic activity.

Only variables that are reported in dollars can be adjusted for inflation. Because the value of a dollar changes over time, inflation adjustments should be made when you want to make comparisons of dollar variables at different points in time. Failing to do so can result in misleading conclusions. For example, if over a 10-year period personal income doubled, but prices also doubled, the real purchasing power of the income has not changed. It’s more important to be more concerned with how much stuff money can buy rather than with the number of dollars involved. Variables that are not adjusted for inflation are called “nominal.” Variables that are adjusted for inflation are called “real.”

Before adjusting variables for inflation, you need to determine which type of deflator or price index to use. Two main types of price indexes are used: (1) one that measures the price of goods and services that people buy like the Consumer Price Index and (2) one that measures the value of goods and services produced like the Price Indexes for Gross Domestic Product.

General Formula:

Real Value = (Nominal Value ÷ Price Index) × 100

Example:

Nominal wage in 1980 = $7.70 
CPI-U in 1980 = 82.4 
Nominal wage in 2000 = $15.16 
CPI-U in 2000 = 172.2

Real 1990 Wage = (7.70 ÷ 82.4) × 100 = $9.34 
Real 2000 Wage = (15.16 ÷ 172.2) × 100 = $8.80

Even though the nominal wage more than doubled from 1980 to 2000, the real wage actually decreased.

The BLS provides an Inflation Calculator. You can plug in any value that you want to convert, along with the years in which you’re interested, and it will tell you what the real value is.

No. For most purposes, the best measure to use is the national Consumer Price Index for All Urban Consumers (CPI-U) (link to it in our database) calculated monthly by the U.S. Bureau of Labor Statistics. The BLS does compute CPIs for a few large metro areas around the county, including Pittsburgh, Cleveland-Akron and Philadelphia, and for the Northeast region of the country in different size classes. (Technically, Erie would fall into the Northeast Urban category, size class B/C.) But those indexes are based on less data than the national CPI-U and therefore are probably not as accurate, so it’s typically better to use the national index for measuring inflation. There is probably not a big difference in the rate of inflation from place to place within the country, also, so using the national rate makes sense.

Adjusting for the Cost of Living (COL) from place to place differs from adjusting for inflation. Inflation adjustments measure changes in prices over time, while the COL from place to place measures the spatial difference in prices at one point in time.

There are a handful of organizations that calculate spatial COL indexes that allow you to estimate the difference in COL for different places. They also offer salary calculators that can tell you what you would need to make in one city in order to maintain the same level of buying power you have in another. Some of the more common sources are:

ACCRA  
DataMasters
Sperling’s Best Places

(When we do this kind of work we use the ACCRA Cost of Living Index data. Note: we are not necessarily recommending any of the others. Caveat emptor!)

Unfortunately, Erie does not currently participate in the ACCRA survey. However, ERIE’s Dr. James Kurre has developed spatial COL estimates for each of Pennsylvania’s 67 counties. These estimates can be used to obtain an understanding of the COL differences within PA. They are available here. 

COL index data can be used to estimate the percentage differences in the COL for various places. For example, if the COL Index value for Boston, MA = 137 and the COL Index value for Buffalo, NY = 99, then moving from Boston to Buffalo would result in a 28% [(137-99)/137] higher COL. That is, it costs about 28% more to live in Boston than in Buffalo. What costs $99 to buy in Buffalo would cost $137 to buy in Boston, according to these data.

The data on the ERIE website are for Erie County unless otherwise indicated. Most of our data are for Erie County, but some are for the City of Erie. The City of Erie is a political unit with very specific boundaries, while data for Erie County are for all units (people, business, etc.) in the county. Since economic activity flows easily across city boundaries (Are you always aware when you’re crossing a city boundary to shop? Does it matter to you?), economists tend to look more at metro areas than cities as the key unit for analysis of local economies. The federal government (actually the Office of Management and Budget) identifies a group of “Metropolitan Statistical Areas” or MSAs which are designed to be “local economies”—primarily labor markets which include the home counties of most of the people who work there, and the workplaces of most of the people who live there. Much more on MSA definitions can be found here. 

The Erie MSA consists of Erie County, so the County and MSA are synonymous.

Erie County Government is the county government for Erie County. Governments in Erie County are those governments operating within the county. For example, Millcreek, North East Township, or the City of Erie governments are governments operating within the county, but none of them is the Erie County Government. This seems obvious, but when looking at government data it is easy to forget that the sum of a variable for all municipalities in Erie County will not be the same as the value of that variable for Erie County government.

The Office of Management and Budget of the federal government defines MSAs. According to their site: “The general concept of a metropolitan or micropolitan statistical area is that of a core area containing a substantial population nucleus, together with adjacent communities having a high degree of economic and social integration with that core.” The most recent definitions were announced June 6, 2003 and are the first to incorporate the idea of “micropolitan” areas. The Erie MSA consists of just Erie County, PA.

About the process of defining MSAs, the OMB says:

“The 2000 standards provide that each CBSA (core based statistical area—metro- and micropolitan statistical areas) must contain at least one urban area of 10,000 or more population. Each metropolitan statistical area must have at least one urbanized area of 50,000 or more inhabitants. Each micropolitan statistical area must have at least one urban cluster of at least 10,000 but less than 50,000 population.

“Under the standards, the county (or counties) in which at least 50 percent of the population resides within urban areas of 10,000 or more population, or that contain at least 5,000 people residing within a single urban area of 10,000 or more population, is identified as a “central county” (counties). Additional “outlying counties” are included in the CBSA if they meet specified requirements of commuting to or from the central counties. Counties or equivalent entities form the geographic “building blocks” for metropolitan and micropolitan statistical areas throughout the United States and Puerto Rico.

If specified criteria are met, a metropolitan statistical area containing a single core with a population of 2.5 million or more may be subdivided to form smaller groupings of counties referred to as “metropolitan divisions.”

As of June 6, 2000, there are 362 metropolitan statistical areas and 560 micropolitan statistical areas in the United States. In addition, there are 8 metropolitan statistical areas and 5 micropolitan statistical areas in Puerto Rico.”

Nearby micropolitan areas include: Bradford, Indiana, Meadville, Oil City, Selinsgrove, and Warren.

Official employment numbers come from the Bureau of Labor Statistics, part of the federal government’s Department of Labor, in cooperation with state departments of labor. The BLS has two different programs that measure monthly employment statistics:

• Current Population Survey or CPS (often called the household survey)
• Current Employment Statistics or CES (often called the establishment or payroll survey)

The Current Population Survey (CPS) is a monthly survey of about 60,000 households nationwide that provides a comprehensive body of data on employment, unemployment, the labor force, and persons not in the labor force. It is a count of people, based on where they live. Monthly CPS data for Erie are available back to January 1990.

The Current Employment Statistics (CES) survey collects data from a sample of over 300,000 businesses each month and provides detailed data by industry category on employment, hours, and earnings of workers on nonfarm payrolls. It is a count of jobs, based on where people work. Monthly CES data for Erie are available back to January 1950.

These two programs have different purposes and techniques, so they generate different measures of employment. In Erie, the CPS numbers have historically been a little greater than the CES numbers, although they have been close recently. While they have their differences, the two surveys show similar trends in employment over the longer term.

The establishment survey (CES) excludes unpaid family workers, domestic workers in private homes, proprietors who own their own businesses, and other self-employed persons—all of whom are covered by the household survey (CPS). Moreover, since the establishment survey is a count of jobs, if a person is employed by two or more establishments, s/he is counted at both, leading to higher totals. The household survey, on the other hand, is a count of people, and would count such a person only once. On the other hand, some people who are on unpaid leave for the entire period are counted as employed under the household survey but not in the establishment survey. Moreover, the CPS counts people based on where they live, while the CES counts them where they work. This can also lead to differences in the totals if there is substantial commuting across county lines. Both programs count both part- and full-time workers it their totals.

The household survey emphasizes the employment status of individuals and provides more detailed information on the demographic characteristics (sex, age, and race) of the labor force at the national level. The establishment survey provides limited information on personal characteristics of workers, but provides data by industry and also data on earnings and hours worked per week. The establishment and household surveys thus can be used together to complement each other.

Which data series should you use? That depends on the purpose for which you need the data. If you are looking for a count of the number of jobs in the area, or need data broken down by industry, or earlier than 1990, the CES is your database. If you want information on the demographic characteristics of the labor force, or you want to include only people who reside in the area, and only need data since 1990, the CPS is the one to use.

The above descriptions are from the U.S. Bureau of Labor Statistics’ “Handbook of Methods.”

It is possible to graph two variables that have different scales by using indexes. Calculating an index allows you to observe movement in a data series, but does not necessarily allow you to observe the actual level of the data. An index will have a base year, which is assigned the value of 100.

To construct an index, you begin with the data point that you want to be the base (usually the first data point you have). The general formula for creating an index is:

Index Value = (data point value / base year data point value) × 100

A hypothetical example is show below to illustrate how to calculate an index.

Step 1: Pick the data point you want to be the base. 
Ex. 1990 Erie Employment = 125,000

Step 2: Calculate the base year value & set at 100. 
Ex. (125,000/125,000) × 100 = 100

Step 3: Calculate index values for remaining data. Use the base year value as the denominator. 
Ex. 1991 Erie Employment = 127,000

1991 Index Value = (127,000/125,000) × 100 = 101.6 
1992 Index Value = (128,000/125,000) × 100 = 102.4

Notice that you can easily determine the percent difference between the base year and any following years. For example, Erie employment in 1991 was 1.6% higher than in 1990 and Erie employment in 1992 was 2.4% higher than in 1990.

The unemployment rate is calculated as the percentage of the labor force who could not find work. The labor force, then, is composed of employed and unemployed persons. To be considered unemployed, people must have actively looked for work, but did not work during the survey period. If people did not work or did not actively look for work, they are not in the labor force and not included its calculation for the unemployment rate.

The Bureau of Labor Statistics answers this question very well and in more detail at its website.

And if you want the really detailed version with all the nitty-gritty, go to the “BLS Handbook of Methods.” Chapter 1 gives information on measuring the national unemployment rate and Chapter 4  presents information on state and local unemployment rates.

An index is a series of values that shows movement or changes in data and not the actual level. Indexes can be used to compare different data series that have different scales (like Erie employment, around 100,000 and U.S. employment around 100,000,000). See the above illustration for how to calculate an index.

SIC (pronounced S-I-C) stands for Standard Industrial Classification, and NAICS (pronounced “nakes”) stands for North American Industrial Classification System. They are both systems for identifying the industry to which an activity belongs. The SIC system is the older system that is being phased out; NAICS the new system that is being phased in. As the name implies, NAICS applies to Canada and Mexico as well as the U.S.; the SIC system was specifically for the U.S. alone. According to the NAICS Website:

The SIC system was developed in the 1930s, and although revised several times since then, it did not accurately reflect the structure of the modern economy. “The SIC system was not based on a consistent economic concept; some industries are demand based while others are production based. NAICS is based on a consistent, economic concept: economic units that use like processes to produce goods or services are grouped together.

“NAICS recognizes the changing and growing services-based economy of the United States and its North American neighbors. NAICS includes 1,170 industries of which 565 are service-based industries. The SIC had 1,004 industries of which 416 were service related industries. Three hundred and fifty eight new industries are recognized in NAICS, 250 of which are services producing industries. There are 20 sectors in NAICS of which 16 are services related.”

The U.S. government is in the process of switching from SIC to NAICS codes, a process which will take several years. As a result, some data are available on one basis, and some on another. Some series will have dramatic breaks due to the changes, and others will simply start or end with the changeover. Unfortunately, there is not a simple correspondence between the two systems, and it is often impossible to splice series across then change period.

In both systems, more digits in the industry number means greater detail.

–List of SIC industries

–List of NAICS industries

When possible, it is probably better to use the NAICS system, since it better reflects the structure of the current economy, and it is the system that will be used by all government agencies in the future.

SOC stands for Standard Occupational Classification system, and it’s analogous to the SIC system explained above except, of course, that it’s for occupations rather than industries. On its Web site, the BLS says:

“The Standard Occupational Classification (SOC) system will be used by all Federal statistical agencies to classify workers into occupational categories for the purpose of collecting, calculating, or disseminating data. All workers are classified into one of over 820 occupations according to their occupational definition. To facilitate classification, occupations are combined to form 23 major groups, 96 minor groups, and 449 broad occupations. Each broad occupation includes detailed occupation(s) requiring similar job duties, skills, education, or experience.” The SOC User Guide is available here and the SOC major groups are here.

You can compare the industrial or occupational distribution for any area by using location quotients. Location Quotients (LQs) are calculated as the percentage of some variable (like employment or income) in an industry for an area (like Erie ) divided by the percentage of that variable in the same industry for the comparison area (like the U.S. ). An LQ greater than 1 indicates that the area has a higher percentage of the variable in that particular industry than the comparison area used in the denominator. An LQ less than 1 indicates that area used has a lower percentage of the variable in that particular industry than the area used in the denominator.

Hypothetical Example:

Total Erie Employment = 130,000
Erie Manufacturing Employment = 27,000 
Erie , Percent Employment in MFG = (27,000/130,000) × 100 = 20.77%

Total U.S. Employment = 130,000,000 
U.S. Manufacturing Employment = 15,000,000 
U.S. , Percent Employment in MFG = (15,000,000/130,000,000) × 100 = 11.54%

Manufacturing LQ for Erie = (20.77/11.54) = 1.8
The manufacturing LQ of 1.8 indicates that Erie has a larger percentage of its employment in manufacturing than does the U.S.

Refer to “How can I compare the industrial or occupational distribution of Erie with that of the U.S. or PA?”

Online Library section of the ERIE website has our own studies of the Erie economy, as well as many others, especially those done for local economic development agencies. You can download them for free from our site. You may also find some of these documents at the Erie County Library, or be able to obtain them from the original authors or sponsoring agency (which are usually given in our listings).

Virtually all of the data we have are posted on the ERIE website. You might try contacting us, though. (Contact Us) If we don’t have the data, we may be able to give you some leads on possible sources. (No guarantees, though!)

We endeavor to present the most recent data, although there is sometimes a lag between the release of the data and when we get it posted to the ERIE site. For most of our data series there is a direct link to the original source of the data and you can check for yourself if there are more recent data or not. Some data series take quite a while for the official agency to compile. For example the local income data typically are not released until about 17 months after the close of the year in question; the 2001 Erie income data were not released until May of 2003. Some very detailed data may only be available through the decennial census, so that the most recent data will refer to the last year ending in zero. On the other hand, some data can be very timely; unemployment rates come out within two months of the date.

Seasonally adjusted refers to data series that have been changed to take into account the regular seasonal patterns that some activities experience. Seasonality occurs when the level of a series is notoriously high or low during certain time periods within a year. For example, employment in retail stores is typically higher during the winter holiday season (e.g., Christmas) and employment in construction is typically lower during winter months. Unemployment rates typically rise each summer when kids get out of school and look for summer jobs, and fall in the autumn when they go back to school. Seasonally adjusted data allow for better comparison across time periods since the effect of any seasonal pattern is taken out.

Seasonality, by definition, can only occur within a year–data that are on an annual basis cannot exhibit seasonal patterns.

Not all data on our site are seasonally adjusted because the original data sources themselves do not offer them. Seasonal adjustment requires relatively sophisticated procedures and adjusted series for most geographical areas below state level are not available.

The Economic Research Institute of Erie is working to seasonally adjusting data for Erie County and they will be available on this website.

If you figure that out, please let us know! But seriously, it is possible to use data on leading indicators to gauge when the economy might turn. Leading indicators are those which typically turn up/down before the economy does.

The Conference Board calculates a leading indicator index for the national economy, which is composed of 10 individual economic variables: building permits, index of consumer expectations, average weekly manufacturing hours, manufacturers’ new orders for consumer goods and materials, vendor performance, interest rate spread, stock prices, average weekly initial claims for unemployment insurance, real money supply, and manufacturers’ new orders for nondefense capital goods. These series typically turn up before the rest of the economy recovers, and turn down before the rest of the economy enters recession. Unfortunately, they are not always accurate indicators of an impending up/downturn. Sometimes the indicators rise or fall and the economy does not follow, and sometimes they give a long lead time and other times a short lead time.

The Conference Board’s leading indicator index is available on their website.

Typically, Erie County has recessionary periods that are longer than the U.S. Erie usually enters a recession before and exits the recession after the U.S.

When a new plant locates in an area and creates employment and income locally, its impact usually generates a larger total effect than that plant’s own impact. This is because of re-spending effects. For example, when an employee receives wages, say $100, for working, they don’t just hold onto it–they spend a fraction of it. Part of the money they spend, say $80 for groceries, then becomes income for the grocer, who then spends part of it with, say, a barber. In this way, the initial $100 of income generates even more local income. It is possible to estimate the total amount that results. If the final total in this example were $290, then the initial $100 would have been subject to a multiplier of 2.9. The multiplier is the amount by which the initial change in income (or employment) in the region is multiplied through the re-spending process to get to the final, total change in income (or employment).

The new company has a multiplier effect because it impacts groups other than the ones it directly interacts with. The money the new company spends is re-spent by those who receive it, including suppliers and employees. The new company’s suppliers spend the money on their suppliers and their employees. The new company’s employees also spend their money on goods and services in the area. This cycle of spending continues and the income that is generated above and beyond the initial amount is the multiplier effect.

See “What is a “multiplier” and how is it interpreted?”

Answer: you can’t predict interest rates. At least, we can’t. Many people try, and some are right some of the time. But they’re also wrong some of the time, and it’s hard to tell whether their next prediction is in the right or the wrong group. You can find some interest rate forecasts on the Economic Forecasts page of the ERIE Web site. 

Watching the actions of the Federal Reserve System may give a clue as to what will happen, since the Fed can increase or decrease the money supply, which affects interest rates. But the Fed doesn’t really set interest rates, they influence them, and there are many other economic actors which influence them too, such as domestic borrowers and lenders, banks and other financial institutions, foreign borrowers and lenders, the government, etc. It’s quite complicated! (Life’s just not as simple as we’d like it to be sometimes, is it?) It may help to know that interest rates tend to be “pro-cyclical”; in other words, they tend to go up during economic upturns and down during economic downturns.

By the way, if you do figure out how to forecast interest rates you can probably make a fortune, because bond prices move in the opposite direction from interest rates—so if you know interest rates are going down, you know that bond prices are going up, so you can buy bonds and wait to reap your profits. Similarly, if you know interest rates are going up, you know bond prices are going down, so you can sell bonds short and make a profit when they go down. Of course, if this were easy to do, there’d be a lot more millionaires around! And accurate interest rate forecasters would just be trading in bonds from a beach somewhere rather than trying to hawk their interest rate forecasts.