Census Count Changes Things
By Marshall J. Vest
EBR and Forecasting Project Director
March 1, 2011
The 2010 Census count came in well below expectations. The low number raises the possibility that population declined during the recession. How does a declining population affect the outlook for recovery? In this quarter’s update, we take a first look using preliminary revisions to population for prior years. We’ve revised our forecast significantly downward.
On December 21, 2010, the U.S. Census Bureau released results for Arizona from the 2010 Census. The Census count shows Arizona’s population at 6,392,017 as of April 1, 2010. That ranks Arizona as the second-fastest growing state during the past decade (up 24.6% or 1.261 million) and the 16th most-populous state. As a result, Arizona gains one seat in the U.S. House of Representatives.
Data for Arizona’s sub-state areas are not yet available. In early February, the Census Bureau began providing summaries of 2010 Census population totals, as well as data on race, Hispanic origin and voting age for multiple geographies within each state. The Census Bureau will deliver the data state-by-state on a rolling basis through the end of March.
In the state “where growth is good and too much is just right,” the count was disappointing as it was more than one-quarter million less than what demographers were expecting. The Census Bureau on February 2nd released their estimate, using the Bureau’s standard estimation procedures, of what the count should have been. That number was 6,654,358, or 262,341 higher than what was actually counted. The difference of 4.1% was by far the largest miss (understatement) for any state. The next-largest shortfall was for Georgia, where the count was 2.0% lower than expected.
Three possible explanations account for the shortfall. Either the estimation methodology broke down, thereby producing estimates that were simply too high. Or, Arizona’s population declined during the recession. Or, a large number of residents were simply not counted. The latter two explanations are most likely, since estimation methods continue to be quite accurate in most states (and in error by only 0.1% nationwide).
Arizona historically has suffered from large undercounts during decennial enumerations and, given Arizona’s series of legislative actions against illegal immigration in recent years, it is understandable why a large portion of the population might not want to be counted. For those same reasons, along with negative characterization in the national press and unprecedented declines in employment, many families no doubt have left the state.
Outflow of people is corroborated by a new report from the Pew Hispanic Research Center1 that the unauthorized immigrant population declined by 160,000 in the three-state area that includes Arizona, Utah and Nevada. The decline is measured from the peak year 2007 through 2010. Other large declines were found in Florida (230,000), New York (200,000), Virginia (100,000), and Colorado (65,000).
So, is it possible that Arizona’s population has actually declined recently – something never before experienced? And if so, how does that change the outlook?
In this quarter’s forecast update, we incorporate a preliminary set of population estimates for the decade spanning 2001-2010 that shows a loss of 40,000 residents in 2010 following an increase of only 27,000 in 20092. Remember that population continues to grow by 45,000 – 50,000 each year due to natural increases of the resident population (births exceed deaths). This means that net migration was negative in both 2009 and 2010 (-25,000 and -85,000, respectively).
A declining population is more consistent with recent large downward movements during the recession in other aggregate measures such as employment, income, and sales. The new path for population alters the dynamic properties of our models; with the new data, the models no longer want to return to the long run growth path as quickly. Given the headwinds facing Arizona’s economy, a delayed and slower pace of recovery makes good sense as well.
In our new forecast, population’s decline of 40,000 (-0.6%) in 2010 is followed by virtually no change in 2011 and a small gain (30,600 or 0.5%) in 2012. Annual growth doesn’t top 100,000 until 2014 and doesn’t exceed a 2% growth rate until 2015. By 2020, the new forecast calls for 7.5 million residents, almost a million less than in last quarter’s forecast. Even so, over a million residents will be added over the decade of the 2010s (Exhibit 1).
Given the loss of mobility of the nation’s population, which is likely to remain a problem for a couple more years, a gain of 1.1 million during the current decade is a reasonable expectation (Exhibit 2).
Since new population counts for metro areas were not available when these forecasts were prepared, we continued to use Census estimates through 2009. Net migration was add-factored downward for 2010 and onward to parallel statewide results. The metro area forecasts are not, however, consistent with the new statewide forecast, so caution should be taken in interpreting the geographic differences.
A scan of the most recent data supports the conclusion that economic recovery in Arizona is still trying to gain traction with little headway so far. Most measures remain near their low points after bouncing along the bottom during the past year. Nonfarm employment in December was up less than one percent during the past twelve months compared to a decline of more than 11% during the recession. Although retail sales registered an encouraging 6% increase year-on-year in November, the level remains in a narrow range and is no higher than at the beginning of 2009.
In short, the economy has yet to break to the upside, and remains close to the bottom of this business cycle.
With mobility restrained by housing economics, we can expect a much slower recovery in employment as well. The economy can recover without the typical boost from construction activity, but the pace will be muted. We expect the economy to create a little less than 20,000 jobs this year and roughly 30,000 next. Both government and construction payrolls will decline both years. Trade, health care, and professional and business services will lead but with small gains. It’s 2015 before the annual increase exceeds 100,000.
Meanwhile, the outlook for the national economy has brightened considerably in the past few months. Credit measures for both business and consumers are starting to improve, banks are beginning to commit their cash to new lending, equity markets are moving up smartly, household net worth is rising, and consumers are boosting spending. The unemployment rate dropped nearly a point in December and January, not because workers were abandoning their job search, but because employment increased. All major sectors of the economy with the exception of housing are now growing. Real GDP is forecast to expand by more than 4% in the first quarter, and 3.2% in 2011.
Renewed growth of the nation’s economy will speed recovery of credit markets, housing prices will stop falling, foreclosures will recede, and mobility of the population will improve. With the passing of time, Arizona’s vacant housing stock will be reduced and home building will resume. Commercial real estate markets will follow housing with a lag of two years or so. By mid-decade, Arizona’s economy should be hitting on all cylinders once again.
2. Estimates provided by ASU’s W.P. Carey School of Business.
For additional information, please contact the Economic and Business Research Center.