As a measure of economic well-being, per capita personal income ain't what it used to be.

• Author: Slaper, Timothy F.

If you don't know how personal income per capita is calculated, you should probably consider using another measure for state or regional standard of living.

Introduction

Per capita personal income (PCPI) is currently (and if one had his way, soon formerly) the standard benchmark for general economic wellbeing at the national, state and regional level. Adjustments for the cost of living across states that are expensive to live in, relative to states with a lower cost of living, often serve to adjust the PCPI metric to make regional comparisons equitable. That said, the PCPI that may have served well as a measure decades ago may no longer be serving us well today or into the future. Rotary phones served our communication needs and the economy plenty well in the 1960s and 1970s, but even computerized touch-tone phones from the 1980s and 1990s cannot keep up with the connection and communication capabilities and needs of today.

We need something more current and more relevant.

Over the years, if not decades, the IBRC has published articles related to PCPI and why Indiana appears to be falling behind in this critical measure and benchmark of Indiana's economic performance relative to other states. The motivation for this article is to alert the reader that the PCPI measure may not be performing as expected.

The motivation for this article is to alert the reader that the PCPI measure may not be performing as expected.

How the measure expresses the relative economic performance across states, regions and counties is an artifact of how the measure is constructed. The economy has undergone considerable restructuring over the last 50 years as we've moved from Meatspace to the Bitworld, and several of the components of PCPI reflect that restructuring. (1) It is unfortunate that the subcomponents of what constitutes PCPI at a regional level are not available on a per resident basis or aggregate basis-for example, county resident dividends earned and capital gains/losses earned, royalty payments from natural resource extraction, returns to intellectual property (such as patents by residents versus corporate patent holders), and property rents associated with land, apartments, homes and other real estate. Such measures would not only help economists understand what drives economic growth-such as jobs or wages and salaries-but would also help us understand the structural income components of economic inequality.

One cannot tackle all these and collateral questions. But one can at least identify points of weakness in current measures and suggest a possible new metric for regional and state economic progress. Spoiler alert: The proposed measure isn't perfect. And it doesn't address the concerns related to making the structural income components of income sources available. One's hope is that it will spur new thinking about what matters in terms of economic statistics and measures of economic performance.

The components of PCPI-the boring stuff-and the devil in the details

In this section, we try to distill the complicated sources and methods of how personal income and PCPI are reckoned in the national product and income accounts. (More detailed information is available directly from the data source for interested readers. (2) )

The sources of personal income inflows and outflows are difficult enough, but add in the geographic dimension-that where people work and earn income can vary greatly from where they live, receive and dispose of income-and the geographically specific identification of PCPI, by state, MSA or county, becomes quite the challenge.

The U.S. Bureau of Economic Analysis provides a lot of interesting detail in how income (as well as product/output) is measured-for example, there are estimates in the national accounts for what the owner of a mobile home would have to pay as rent if he or she rented that housing unit. And those imputed rent estimates are part and parcel of total personal income. But we will gloss over imputed for the moment to focus on the major income categories, with reference to their sub-components as warranted. The BEA line numbers relate to tables and data sets available on the BEA website.

* Headline personal income (BEA Line 10)

* Wages and salaries (Line 50)

* Net earnings by place of residence (Line 45)

* Dividends, interest and rent (Line 46)

* Personal current transfer receipts (Line 47)

* Adjustment for residence (Line 42)

Adjustment for place of residence

Doing the math on the national level, net earnings by place of work

(Line 35) is wages and salaries plus employer contributions (61 and 62) plus proprietors' income (70). These proportions/values and the additivity vary by region. This is why there is this an income and earnings adjustment for residence (in contrast to place of work). At the

national level, the calls and the puts even out and, in round numbers of billions of dollars, cancel each other out. On a state, regional and especially county level, adjustment for residence can matter greatly.

Let's say that about 40% of the workers at Crane NSWC in Martin County, Indiana, live in Monroe County. Crane is a commuter campus in other words. While those paychecks are cut in Martin County, a vast majority of the Crane NSWC workforce lives outside the county. And yet, the Social Security payments and the benefits, be they matching retirement accounts or health insurance benefits, are registered from Martin County and those benefits are personal income. The adjustment for place of residence allocates the earnings and supplements to where the workers live. After all, the PCPI uses per person as the denominator and the relevant number of persons are residents based on census counts.

These adjustments can matter in terms of scale depending on population of the county and the relative size of the employer. The Martin County adjustment is -1.3 and the earnings by place of work are almost 2.1 times greater than the headline personal income for the county. In contrast, the Monroe County adjustment is -0.04 and the earnings by place of work are 0.73 times greater--really means less-than the headline personal income number.

The four counties with the largest adjustment in 1969 were counties with a small residential footprint: Martin, Indiana; Barrow-North Slope Division, Alaska; Butte, Idaho; and Nye, Nevada. In 2019, Butte, Idaho, and Martin, Indiana--both with federal facilities--were still in the top 10 commuter counties, with other counties popping up because of industry investments distinct from residential communities, for example, Storey County, Nevada. Constrained and dense city centers also show this same effect, for example, Washington, D.C., where for decades there has been the threat of a commuter tax, and the county of

San Francisco, while not registering as highly on the commuter factor as Washington, D.C., is another example of why the adjustment for residence is required. The commuter effect is also evident in smaller cities and places in which there is a solitary large employer that serves as an employment magnet to the surrounding region; Howard County --ranking 238th of 3,100 plus counties in terms of the adjustment for residence--is another example of a cluster of production activity drawing commuters from well beyond county boundaries.

Types of income

Having stated the need to adjust geographically incoming income differently than geographically outgoing income, we turn our attention to three of the more important income source categories for our analysis: 1) net earnings by place of residence; 2) dividends, interest and rent; and 3) personal current transfer receipts.

There are plenty more subcategories and dimensions and wrinkles in the data source and estimation method, but rather than getting into the thick and tall weeds, let's just see the degree to which these three (plus wages and salaries) can tell a story about economic...