In a report released on March 23, AEDI presented a preliminary forecast for the Arkansas economy, taking into account the impact of the novel coronavirus disease 2019 (COVID-19). In addition to disruption of world supply chains, the decline in consumer spending associated with “social distancing” is generating forecasts with significant declines in aggregate demand, creating the conditions for a severe recession.
In this note we update that forecast with new estimates of the magnitude of the downturn, taking into account the impact of the Coronavirus Aid, Relief, and Economic Security (CARES) Act into law on March 27. We also update and extend our previous guidance on how the forecast is likely to impact sales tax receipts of local governments.
The forecast is based on the latest update to the national outlook from IHS Markit, dated April 2, 2020. The new forecast sees GDP growth falling at a 26.5% annual rate in the second quarter – down from a 12.6% rate of decline forecast just two weeks earlier. National unemployment is now projected to peak at over 10% in the fourth quarter of the year. According to IHS, the new forecast “reflects inclusion of new high-frequency data and reports on developments in industries directly affected by social distancing, as well as new judgment on how efforts to slow the spread of coronavirus disease 2019 (COVID-19) will permeate the economy.”
As with our previous forecast for the Arkansas economy, we have applied the recent IHS forecast adjustments to our model for Arkansas, adjusting the components of each aggregate separately in order to capture the unique characteristics of the Arkansas economy. Our state forecast is also informed by an update to the IHS Arkansas model published on March 27 that reflected the March 20 national forecast assumptions.
In a nutshell, the latest forecast projects far sharper downturns in employment and consumer spending, with deeper and broader effects. Nonfarm payrolls are expected to decline by over 100,000 and unemployment is expected to peak at 9.2%. Consumer spending is expected to be down over 9% in the second quarter. Most components of income are also expected to decline more sharply than in the previous forecast, but estimates of transfer payments stemming from the CARES Act are expected to buffer the impact on total personal income.
The following sections present our latest post-COVID-19 forecasts for Arkansas, comparing the projections to our forecast from March 30, as well as the previous, pre-COVID-19 baseline forecast.
The latest forecast indicates a decline of 108,000 jobs by the fourth quarter of 2020, an 8.4% drop. The previous forecast had indicated a loss of 77,000 jobs.
As shown in Table 1, job losses are concentrated in service-providing sectors that are considered especially impacted by “social distancing.” However, the new forecast shows declines across the entire economy. Even Health Care & Social Assistance is now expected to see job losses by the end of the year. In light of new patterns of consumer behavior, some sectors have been revised to show smaller declines than previously forecasted. For example, employment in Accommodation and Food Services was previously expected to decline by 13%, but drive-through, carry-out, and delivery options are now anticipated to limit that sector’s decline to only 9%. Employment in nondurable goods production—primarily in food processing—is expected to decline by only 2.9%, as opposed to a previously-forecast decline of 3.5%. The largest downward revisions are associated with employment in sectors that are not directly impacted by COVID-19 or social-distancing practices, as the effects of the economic downturn spread throughout the economy.
The Arkansas unemployment rate was previously forecast to peak at 8.7%. The latest forecast suggests 9.2%. As dire as that sounds, the forecast for Arkansas is lower than the national projection of 10.3%. The increase in forecasted unemployment is buffered somewhat, particularly in this most recent forecast, by a decline in the labor force participation rate. The participation rate is expected to fall by over a full percentage point, from 56.8% to 55.5%.
The latest forecast for personal income includes a surge in transfer payments in the second quarter of 2020, reflecting payments associated with automatic stabilizers such as unemployment insurance, as well as the large cash transfers approved under the CARES Act. In the U.S. forecasts, IHS Markit predicts an increase of nearly 24% in second-quarter transfers, with an additional 3% increase in the third quarter. That translates to a path for total personal income that is basically flat through the third quarter (down 0.2% from Q1).
If we apply the national adjustment factors directly to personal income, Arkansas follows a similar trajectory—labeled as “U.S. Equivalent” in Figure 3. However, transfer payments account for a larger share of total personal income in Arkansas than for the U.S. average (about 24% vs. 17%). Consequently, if transfer payments surge by as much as 24% in the second quarter, Arkansas would see an increase in total income through the third quarter – up 2.6% from Q1 (labeled as “April Forecast” in Figure 3).
Is it reasonable to expect transfer payments to increase in the same proportions as the national total? On the one hand, a given dollar-amount transfer payment translates to a smaller percentage increase when the base is larger, so we might expect something smaller than a 24% increase in Arkansas. On the other hand, as a relatively low-income state we are likely to have fewer people excluded from transfer programs by means-testing. Looking back to the stimulus programs of the “great recession,” transfer payments in the second quarter of 2008 were up from the previous year by 25.6% (U.S.) and 25.3% (Arkansas). So it is, perhaps, not unreasonable to expect similar growth rates of transfer payments in the present situation. All of this discussion highlights the fact that there is considerable uncertainty about all forecasts, particularly in unusual times and circumstances.
Overall, the new personal income forecasts indicate sharper declines in all categories of earnings and other income, but higher transfer payments will significantly offset those losses, at least for a time.
Despite the boost to incomes from expected transfer payments, the balance of forces is now expected to result in sharper declines in Personal Consumption Expenditures (PCE) than in the previous forecast. Spending in the second quarter is now projected to decline by 9%, with additional weakness expected in the first quarter statistics as well (-0.7%). The trajectory for recovery is now more optimistic, however, with growth beginning to rebound in the third quarter.
Table 2 shows the initial declines expected for components of consumer spending, comparing the previous forecast to the current outlook. As is typically the case during recessions, durable goods purchases are expected to fall dramatically, with declines in purchase of motor vehicles down 33%. Nondurable Goods purchases are less-affected during economic declines, and particularly in the present situation with meals-at-home substituting for dining-out, overall Nondurable Goods purchases are expected to decline by less than 7%. Nondurables also includes gasoline, where price reductions account for much of a 42% drop in spending, as well as Food and Beverages Purchased for Off-premises Consumption, which are expected to increase by 15%. Spending on services also tends to be more resilient to economic downturns, but some sectors in particular are expected to show sharp declines. While overall spending on services is projected to decline less than 9%, spending on Recreation is expected to drop by 40% and Accommodations and Food Services by over 60%.
Figure 5 shows index values (relative to 2019:Q4=100), illustrating the relative paths for total PCE, Durable Goods, Nondurable Goods (less Gasoline & Other Energy) and Services. The relative severity of the downturn in Durable Goods is evident. With the exclusion of gasoline, Nondurable Goods are projected to show a modest decline in the second quarter and remain relatively stable. However, the dashed line shows that the stability of Nondurables is attributable to Food and Beverages Purchased for Off-Premises Consumption.
Implications for Local Sales Tax Collections
In a follow-up to our forecast from two weeks ago, we prepared a report on how changes in consumer spending are likely to affect local sales tax collections. This section presents an update on that report.
For assessing the impact on county and municipal sales tax collections, the question is how the tax base is reflected in the various components of spending. Durable goods are taxable and represent a significant share of the sales tax base. Nondurable goods, excluding gasoline, are also taxable. Services are generally not taxable, with the exception of some categories that are expected to be hard-hit by the downturn; in particular, Recreation and Food Services & Accommodation. On the other hand, the surge in grocery sales represent a component of county and municipal sales tax bases that should help offset declines in other categories.
As a proxy for local sales tax bases in Arkansas, we take a simple sum of Durable Goods plus Nondurable Goods (less Gasoline and other energy) plus two services components, Recreation and Food Services & Accommodation. This isn’t a perfect measure. For example, it excludes the utilities component of Housing and Utilities, as well as some taxable components of Other Services. Moreover, as a measure of consumer spending, it does not include some taxable business-to-business sales. Nevertheless, this aggregate measures some of the key components of the tax base.
Figure 6 shows the trajectory of this measure compared to total Personal Consumption Expenditures. Figure 6 also compares the current outlook to our previous forecast, in which grocery sales were expected to buffer the decline in other categories of spending. In the new forecast, the steepness of declines in spending on Durable Goods, Recreation, and Food Services & Accommodation overwhelms the grocery effect. The proxy for taxable sales was projected to decline by 4% from 2019:Q4 through 2020:Q2 in the March forecast. Now it is expected to fall by 18% and it is not expected to recover to 2019 levels until late in 2021.
Figure 7 breaks down the outlook for the taxable sales proxy in two ways: as year-over-year growth rates and as deviations from trend. It is hoped that these figures might help local governments anticipate the timing and magnitude of declines in taxable sales and tax revenue.
In terms of year-over-year growth rates, the first quarter of 2020 is now expected to have declined by nearly 3% from the previous year. The second quarter year-over-year growth rate drops to -16.8%. The figures for deviations from trend illustrate the lasting impact of reduced revenues relative to a pre-existing outlook, dropping to 18% below-trend and remaining in negative territory through the end of 2021.
These estimates suggest large declines in sales beginning in the second quarter. There is typically a one-month lag in the remittance of sales tax proceeds by the vendors and an additional month before the revenue shows up in local government accounts. Hence, the hit to sales tax receipts might not show up in full impact until near the middle of calendar year 2020.
These comparisons are subject to a great deal of uncertainty. They do not reflect seasonal patterns in the data and do not reflect regional differences in consumption patterns and growth trends. As with any forecast, they should be used only as rough estimates.