Key Points
Government spending in 2020 did not significantly increase consumption or investment relative to 2019, and, instead, accrued mostly to U.S. household savings which rose from $1.2 trillion in 2019 to $4.8 trillion (annualized and seasonally adjusted) in 2020 Q2.
As the economy reallocated resources in response to the pandemic, many sectors saw production and consumption near capacity. The total consumption of durable goods did not decline in 2020 and in fact increased by 18 percent from 2019 Q1 to 2020 Q3. Though nondurable goods production decreased by 5.4 percent in 2020 Q2 relative to 2019 Q1, it had grown by 3 percent in 2020 Q3 relative to 2019 Q1. Imports of nondurable consumer goods except food grew by 3.5 percent and imports of foods, feeds, and beverages grew by 6.8 percent in 2020 Q3, implying demand in excess of domestic supply.
Using the implied marginal propensities to consume by household in the PWBM dynamic model, we estimate that 73 percent of the proposed $1400 relief payments in Biden’s COVID relief plan will be directed to household savings.
Background: Marginal Propensities to Consume in the 2021 Economy
Introduction
This brief provides technical background to our macroeconomic analysis of the 2021 Biden relief plan.
The Biden administration has recently proposed a one-time $1400 per-person payment as part of Covid-19 recession relief. The rationale given for these payments is (1) to support struggling families and (2) to spur economic growth. Relief transfer payments, by definition, provide funds which a household can use to spend on consumption, and transfer payments are redistributive regardless of whether households spend or save. However, the broad nature of the proposed policy implies that poorer and unemployed households do not receive more aid than many relatively well-off households.
The macroeconomic effects of the relief depend on (a) the propensity of recipients to consume rather than save and (b) the response of the production sector to any increases in consumption. Transfer payments during a recession can have the following effects:
Stimulate the economy. If the relief boosts aggregate demand which is, in turn, matched by domestic production increases. In this case, the transfer payments overcome institutional failures or frictions which prevent households from borrowing to temporarily finance consumption.
Support domestic consumption without an impact on domestic production. Aggregate demand increases are filled by foreign producers; the trade deficit increases as foreigners accept U.S. assets (future promises to pay) in return for providing current consumption. This case may occur when domestic production cannot expand in the short run, perhaps due to insufficient productive capacity or other restrictions.
Reallocate resources into asset markets. If a large portion of the relief is saved, then the policy produces a national portfolio reallocation: the government borrows to finance the transfers, but the borrowing proceeds are given to private portfolios which are allocated to private capital markets and government debt.1 Government borrowing thus redirected to the capital markets can have the effect of increasing real capital formation and thus growing the economy. If, on the other hand, redirected relief money increases the price of capital assets (relative to wages) without generating new real investment, then the policy may produce unintended wealth redistribution in the economy.
In the Covid-19 recession, all three effects likely occur in some measure. Increased risk aversion may suppress some economic activity and this "output gap" can be ameliorated by (1). Slowness of investment and production to adjust to major shifts in consumption allocation may be described by (2). Transfers to a large portion of the population rather than relief to only the hard-hit households are likely generating (3), and the extent of recent speculative asset price increases has raised concerns about a financial bubble, especially as the financial industry has recently popularized zero-commission and fractional share trading which eases the purchase of capital assets in smaller amounts.
To understand the potential effect of proposed relief payments on short-run consumption and saving, we look at the marginal propensities of consumption for households in the PWBM dynamic overlapping generations model. From aggregate consumption and production data in relation to the relief payments provided in 2020, we find that the Covid-19 economic shock had very uneven impacts on the economy. We, therefore, implement a targeted shock in our model. Our full analysis focuses on long-run macroeconomic consequences, here we look only at the short-run effect. Relief payments given to households are projected to be spent and saved according to the propensities in the model. Our analysis suggests that effect (3) dominates in the current environment with 73 percent of the relief payments likely to be allocated to the asset markets.2
Production and Consumption in the Covid-19 recession
Using national aggregate data from the Bureau of Economic Analysis (BEA), we measure percent changes in sectoral production and consumption since the first quarter of 2019. The nominal amounts are deflated by BEA’s sectoral price deflators in order to estimate real changes. We use this data to understand the nature of the recessionary shock and the potential consequences of broad transfer payments.
Figure 1 shows changes in aggregates—production (“value added”), consumption, price (both since 2019 Q1 and on an annualized quarter-on-quarter basis) and imports—for durable goods, nondurable goods, and services.
Total consumption and production in the U.S. economy declined by 1.3 percent and 1.9 percent, respectively, from 2019 Q1 to 2020 Q3. Declines in consumption occurred primarily for services (6.4 percent from 2019 Q1 to 2020 Q3) which were most affected by lockdowns and decreased in-person contact.
Figure 2 shows changes in aggregates—production (“value added”), consumption, price (both since 2019 Q1 and on an annualized quarter-on-quarter basis) and imports—for durable goods, broken down by sector.
The total consumption of durable goods did not decline in 2020 and increased considerably in 2020 Q3 by 18 percent compared to 2019 Q1. It was most notably driven by an increase in demand for recreational goods and vehicles as people sought entertainment which was less restricted by stay-at-home orders and lockdowns. In the meantime, real value added by durable goods-producing manufacturing industries declined by 14 percent in 2020 Q2 and did not recover fully in the third quarter. Imports of consumer durable goods except automotives increased by 2.1 percent in the third quarter relative to 2019 Q1 while imports of automotive vehicles, engines and parts decreased by 10.7 percent. Since demand for durable goods went up, the mismatch between supply and demand was reflected in an annualized 8.2 percent increase in the price index for durable goods in 2020 Q3 from the previous quarter. We can suppose that U.S. manufacturers are still adjusting to demand shifts, supply chain disruptions, and Covid-19 related production restrictions. These observations suggest that durable goods production is near full capacity and that further increases in aggregate demand would primarily increase price inflation and imports.
Figure 3 shows changes in aggregates—production, consumption, price changes, and imports—for nondurable goods, broken down by sector.
The change in nondurable goods consumption was more uneven, with a decline of over 20 percent from 2019 Q1 to 2020 Q2 in expenditures on clothing and footwear as well as gasoline. The former recovered in the third quarter while the latter was still down by 10 percent in 2020 Q3 as many people traveled less and worked from home. Compared to durable goods, nondurable goods manufacturing industries experienced a modest drop in production in 2020 Q2 except for textile product industries—total value added decreased by 5.4 percent during the second quarter followed by an increase of 3 percent in the third quarter relative to 2019 Q1. Imports of nondurable goods except food grew by 3.5 percent in 2020 Q3 and imports of foods, feeds and beverages grew by 6.8 percent. The modest rise in the aggregate price index for nondurable goods reflects less of a mismatch in supply and demand. Since consumption of nondurable goods is fairly stable and has mostly recovered to pre-pandemic levels, we do not expect transfer payments to further stimulate demand for these goods.
Figure 4 shows changes in aggregates—production, consumption, price changes, and imports—for services, broken down by sector.
The overall increase in goods consumption contrasts with a significant reduction in expenditures on most service industries apart from financial services and insurance, which do not require in-person contact, and housing and utilities, which were not disrupted by the pandemic.3 Most service sectors saw a sharp decline in the second quarter of 2020, for example, by 19.3 percent for health care, 38.5 percent for transportation services, 38.3 percent for food services and accommodations, and more than 50 percent for recreation services. Expenditures on affected sectors were still noticeably below their pre-pandemic levels in the third quarter. Production declines in these sectors track consumption since consumers cannot spend their transfer payments on these sectors due to the shutdowns of restaurants, bars, movie theaters and gyms, and other types of businesses where services were provided. We observe a modest increase in the production of retail and information service industries in 2020 Q3, perhaps fueled by shifted consumer demand towards online services. Overall, expenditures on services were still 6.4 percent below their 2019 Q1 level while production was lower by 2.4 percent.4 Since the decline in consumption and production of these affected industries was due to pandemic-related restrictions, we do not expect transfer payments to increase demand and stimulate production until the pandemic is under control and restrictions are lifted.
From the above analysis alone, we cannot rule out the possibility that some portion of stimulus payments in 2020 restored aggregate demand and therefore production. However, neither can we rule out that the recovery was caused by other policies implemented at the time or by declines in risk-aversion as households acquired more information about the pandemic. Nevertheless, the recovery in consumption and production in 2020 Q3 (and assuming the trend continues into 2021) implies that those sectors of the economy have little “output gap” to restore through additional stimulus.
Figure 5 shows net saving by sector as well as aggregate domestic investment.
Household net saving went up significantly in the second and third quarters of 2020. This large flow results directly from national account identities because the deficit of one sector (the government in this case) must be the surplus of at least one other sector. After looking at production, consumption, investment, and trade flows, private sector savings remains as the source of government borrowing. In 2019, private household savings was $1.2 trillion while the government had a budget deficit of $1.3 trillion. On a seasonally adjusted annualized basis, the first quarter of 2020 began with household savings of $1.6 trillion and a budget deficit of $1.4 trillion. Because of its policy response to the pandemic, the government budget deficit jumped to $5.2 trillion in 2020 Q2 and $3.6 trillion in Q3 (seasonally adjusted and annualized). These monies increased household savings to $4.8 trillion in 2020 Q2 and $2.8 trillion in Q3 (seasonally adjusted and annualized). Some of the government’s deficit was matched by increased lending from foreigners and the business sector, but these surpluses are small compared to U.S. household savings.5 The large increase in savings did not manifest in increased real investment: gross investment was $4.5 trillion in 2019, dropped to $3.9 trillion in 2020 Q2, and recovered to $4.5 trillion in 2020 Q3 (seasonally adjusted and annualized).
Once the pandemic ends and restricted sectors recover, households will be able to spend on consumption in those sectors. This increased consumption is referred to as ‘pent-up demand’ or as deferred consumption. However, it is not clear that significant amounts of recessionary savings are used toward consumption recovery. A 2021 study finds that post-recession recoveries of nondurables and services tend to be muted and that it is primarily durable goods which follow a “pent-up demand” cycle.6 This effect is consistent with a permanent-income theory of household finances where households smooth consumption through time, based on expected long-run income. Buying durables, as a household asset, is then a form of portfolio reallocation. Since the Covid-19 recession is primarily in service sectors, we do not expect to see significant flows from household savings into consumption of recovering sectors. Rather, the likely outcome is the resumption of pre-recession spending patterns from current income (with some temporary increase from near-term deferral).7
Marginal propensity to consume
The portion of the relief payment which is spent on consumption rather than put into savings depends on a variety of factors affecting the household. If income earners in the household are employed in a sector of the economy which is hard-hit by the recession and suffer significant income loss, then it is likely the transfer payment would be spent to compensate for the income loss. Lifecycle income planning by the household determines how the current shock affects saving. Wealth and age determine the degree to which transfers are spent on consumption with older and poorer households more likely to spend rather than save. Other factors may also influence the consumption-savings decision, including perceived returns in the asset markets.
Empirical estimates of the aggregate marginal propensity to consume (MPC) in the U.S. range from 0.05 to 0.9 depending on the event and sample of the study.8 A study using recent data at the household level estimates the MPC by wealth quintile; they measure 0.218 for the first quintile, and 0.166, 0.002, 0.002, 0.015, for the subsequent wealth quintiles.9 Another work surveying households on their spending of the 2020 relief payments finds limited effects on consumption.10
The consumption propensities of households in the PWBM dynamic overlapping generations model arise from the realized responses of model agents’ optimizing behavior. After calibrating for the current economic environment, the values for the MPC by income quintile are 0.55, 0.40, 0.22, 0.13, and 0.12 respectively. These values are higher than the empirical estimates reported above since we include the effect of the ongoing recession in our analysis. During recessions, people use extra income to smooth out consumption and therefore MPCs tend to be higher than in normal times.
Using these values, we calculate the short-run effect on allocations between spending and saving in the model which result from the proposed relief payments. The model projects that, in aggregate, 27 percent of the relief transfer goes toward increased consumption and 73 percent goes to household savings.
Conclusion
Poorer households exposed to severe income shocks, due to their employment in hard-hit sectors, are supported by relief payments. Transfer payments, as redistribution, can reallocate economic activity by increasing consumption relative to savings and this demand growth, and in turn, could produce additional investment in order to increase production capacity. However, in a recession, firms understand that once a recovery occurs, the transfer payments will stop and there is no need to increase production capacity beyond pre-recession levels. As such, transfer payments serve to support production sector income (and reduce risk of owning productive assets) but do not stimulate additional investment. When provided to affected households, relief transfer payments support their consumption and restore income to production sectors which serve them.
Transfer payments to non-affected households, on the other hand, may produce consumption and savings beyond pre-recession levels. According to PWBM’s analysis, the Biden administration’s proposed $1400 relief payments, due to their distribution to a large section of the U.S. population, are projected to result mostly in additional savings and thereby cause a $348 billion allocation from government borrowing to household portfolios. National accounts data appear to indicate that most economic sectors are operating near pre-recession levels, with notable exceptions for service sectors reliant on in person interaction and transportation. Since production and consumption in these underperforming sectors are limited while the pandemic continues, stimulus payments are unlikely to expand economic activity in those sectors.
This analysis was conducted by Marcos Dinerstein, Zheli He, and Xiaoyue Sun and directed by Efraim Berkovich. Prepared for the website by Mariko Paulson.
-
If households save into government bonds, then the net effect of that portion of the transfer payment superficially appears to be neutral—that is, like borrowing a dollar from yourself. However, the distribution effects may not be neutral since government debt is repaid by all taxpayers while only some acquire government bonds from the transfer payment. ↩
-
Paying down household debt is equivalent to savings when one considers liability as a ‘negative’ asset. ↩
-
Rent forbearance and moratoriums reduced evictions of households experiencing income shocks. ↩
-
As most of the services consumed by households are produced domestically, trade in services does not play an important role. ↩
-
On a seasonally adjusted annualized basis, net lending from foreigners increased from $0.43 billion in 2020 Q1 to $0.57 billion and $0.7 billion in the second and third quarters of 2020, while net private saving by domestic business decreased from $0.4 billion in 2020 Q1 to $0.2 billion in the second quarter and went back up to $0.67 billion in the third quarter. ↩
-
Martin Beraja and Christian K. Wolf. (2021) ”Demand Composition and the Strength of Recoveries”, NBER Working Paper ↩
-
By ’near-term’ deferral, we mean, for example, taking a vacation in autumn 2021 which was scheduled for spring 2021. Vacations not taken in 2020 are lost consumption. ↩
-
Carroll, C., Slacalek, J., Tokuoka, K. and White, M.N. (2017), The distribution of wealth and the marginal propensity to consume. Quantitative Economics, 8: 977-1020. ↩
-
Fisher, Jonathan, David Johnson, Timothy Smeeding, and Jeffrey P. Thompson. 2019. “Estimating the Marginal Propensity to Consume Using the Distributions of Income, Consumption, and Wealth.” Federal Reserve Bank of Boston Research Department Working Papers No. 19-4. ↩
-
“HOW DID U.S. CONSUMERS USE THEIR STIMULUS PAYMENTS?” by Olivier Coibion, Yuriy Gorodnichenko, and Michael Weber, 2020 ↩
,Durable goods,Nondurable goods,Services,Total 2019 Q1,0,0,0,0 2019 Q2,3.037886283,1.285902325,0.48241206,0.913451257 2019 Q3,4.626665891,2.058793749,0.993201301,1.589374637 2019 Q4,5.435604958,1.866414661,1.506355306,1.989582379 2020 Q1,1.961240761,3.634952243,-1.090156666,0.192466318 2020 Q2,1.530582553,-0.506260758,-13.60449305,-9.416338252 2020 Q3,18.03526742,6.466637416,-6.360035471,-1.286927565
,Durable goods,Nondurable goods,Services,Total 2019 Q1,0,0,0,0 2019 Q2,-0.113397052,1.52023485,0.305004,0.37097038 2019 Q3,-0.089097683,4.780876494,0.924185,1.010010396 2019 Q4,0.55078568,4.246173202,1.672553,1.602613151 2020 Q1,-0.95577515,2.36946949,0.123072,0.31925616 2020 Q2,-14.05313462,-5.378486056,-9.43754,-8.695376854 2020 Q3,-0.40498947,2.998532187,-2.36741,-1.866988913
,Durable goods,Nondurable goods,Services,Total 2019 Q1,0,0,0,0 2019 Q2,-0.248375783,0.791243571,0.712662407,0.627281231 2019 Q3,-0.692232507,0.669513791,1.328416449,0.97760496 2019 Q4,-1.369516472,1.109769829,1.86612924,1.363694723 2020 Q1,-1.792675214,0.978910316,2.445006646,1.687423195 2020 Q2,-2.573449089,-0.782113838,2.534196647,1.277489408 2020 Q3,-0.626688898,0.18665233,3.237425496,2.19089892
,Durable goods,Nondurable goods,Services,Total 2019 Q1,0,0,0,0 2019 Q2,-0.989807827,3.202736808,2.88126793,2.532832713 2019 Q3,-1.768003341,-0.48222217,2.468107307,1.399848625 2019 Q4,-2.700238922,1.76082111,2.139609557,1.538201448 2020 Q1,-1.70512508,-0.5166887,2.292540356,1.28362578 2020 Q2,-3.142381056,-6.795439841,0.348700424,-1.602800318 2020 Q3,8.235500789,3.963185988,2.771744964,3.656650171
,Motor vehicles and parts,Furnishings and durable household equipment,Recreational goods and vehicles,Other durable goods 2019 Q1,0,0,0,0 2019 Q2,2.689204764,2.403003755,5.075322812,0.97541943 2019 Q3,2.785247791,3.729662078,8.662123386,2.809207959 2019 Q4,3.572800615,4.455569462,9.93543759,2.887241514 2020 Q1,-4.744525547,3.429286608,11.29842181,-1.989855638 2020 Q2,-4.475605071,1.301627034,20.87517934,-19.82052282 2020 Q3,10.27660392,15.36921151,34.23601148,10.80764729
,Wood products,Nonmetallic mineral products,Primary metals,Fabricated metal products,Machinery,Computer and electronic products,Electrical equipment_appliances_and components,Motor vehicles_bodies and trailers_and parts,Other transportation equipment,Furniture and related products,Miscellaneous manufacturing 2019 Q1,0,0,0,0,0,0,0,0,0,0,0 2019 Q2,0.666666667,0,4.985754986,-1.126760563,-1.654676259,0.356400356,0.505050505,0.619834711,-1.707156927,-2.5,0.569476082 2019 Q3,5,1.214574899,18.94586895,-1.61971831,-2.661870504,-1.395901396,2.188552189,0.619834711,-3.020354563,-4.642857143,0.227790433 2019 Q4,7,0.4048583,28.06267806,-2.535211268,-3.884892086,1.455301455,1.851851852,0,-2.823374918,-6.071428571,0.113895216 2020 Q1,6.666666667,-1.417004049,36.89458689,-4.084507042,-6.258992806,0.594000594,2.525252525,-3.925619835,-7.025607354,-6.428571429,-0.227790433 2020 Q2,-6,-5.060728745,23.36182336,-19.08450704,-16.90647482,-1.217701218,-7.575757576,-43.04407713,-30.00656599,-16.07142857,2.164009112 2020 Q3,6.666666667,1.821862348,42.87749288,-11.54929577,-10,3.712503713,-4.377104377,14.18732782,-24.35981615,-7.142857143,15.261959
,Motor vehicles and parts,Furnishings and durable household equipment,Recreational goods and vehicles,Other durable goods 2019 Q1,0,0,0,0 2019 Q2,0.516967805,-0.220478198,-1.728734039,0.796240238 2019 Q3,0.583508215,-0.260461291,-3.037613627,0.240285516 2019 Q4,0.278446025,-1.220055519,-4.029269853,-0.207305151 2020 Q1,0.03685315,-0.701417686,-5.166209048,-1.106020094 2020 Q2,-0.234426985,-0.655722723,-6.797643484,-2.545377448 2020 Q3,4.315913395,1.020140055,-6.439101112,-2.62076114
,Motor vehicles and parts,Furnishings and durable household equipment,Recreational goods and vehicles,Other durable goods 2019 Q1,0,0,0,0 2019 Q2,2.083961898,-0.879000438,-6.737682485,3.223203193 2019 Q3,0.265055797,-0.160189449,-5.222122464,-2.188065551 2019 Q4,-1.207661786,-3.793217675,-4.028559454,-1.774143881 2020 Q1,-0.960211133,2.116772851,-4.655147538,-3.553956228 2020 Q2,-1.080316436,0.184198056,-6.705696716,-5.69594774 2020 Q3,19.53069789,6.920368417,1.547671644,-0.309051608
,Durable Goods_Except Automotive,Automotive Vehicles_Engines_and Parts 2019 Q1,0,0 2019 Q2,-3.672131148,1.050788091 2019 Q3,-1.464480874,0.325243933 2019 Q4,-6.207650273,-5.604203152 2020 Q1,-15.45355191,-8.081060796 2020 Q2,-25.15846995,-57.29296973 2020 Q3,2.142076503,-10.68301226
,Food and beverages purchased for off-premises consumption,Clothing and footwear,Gasoline and other energy goods,Other nondurable goods 2019 Q1,0,0,0,0 2019 Q2,1.062293729,2.702032722,0.022391402,1.356623515 2019 Q3,2.268976898,2.18145761,-0.403045231,2.544777443 2019 Q4,1.804867987,3.61923649,-1.119570085,2.190104628 2020 Q1,8.900577558,-6.817055032,-5.665024631,5.18708991 2020 Q2,7.353547855,-21.14526525,-23.42140618,5.231424011 2020 Q3,8.807755776,2.255825483,-10.27765338,10.79092038
,Food and beverage and tobacco products,Textile mills and textile product mills,Apparel and leather and allied products,Paper products,Printing and related support activities,Petroleum and coal products,Chemical products,Plastics and rubber products 2019 Q1,0,0,0,0,0,0,0,0 2019 Q2,1.045150502,-1.242236025,-2.43902439,0.764818356,0.272479564,5.411550705,0.759013283,1.466666667 2019 Q3,3.051839465,-1.242236025,-2.43902439,3.824091778,1.36239782,14.91587085,3.510436433,2.266666667 2019 Q4,0.919732441,-2.48447205,-3.658536585,2.868068834,1.089918256,13.36971351,4.237824162,2.933333333 2020 Q1,0.877926421,-3.726708075,-4.87804878,3.824091778,0,-0.45475216,4.648956357,2.666666667 2020 Q2,-12.41638796,-23.60248447,-17.07317073,0.191204589,-15.80381471,-9.049567985,4.301075269,-10.8 2020 Q3,1.797658863,-10.55900621,2.43902439,6.118546845,-2.179836512,2.091859936,6.072106262,1.733333333
,Food and beverages purchased for off-premises consumption,Clothing and footwear,Gasoline and other energy goods,Other nondurable goods 2019 Q1,0,0,0,0 2019 Q2,0.092964415,-1.328520732,7.127594628,0.394877977 2019 Q3,0.05079499,-0.512602748,4.34010434,0.596113869 2019 Q4,0.242474195,-2.08579777,6.956931957,1.353596142 2020 Q1,1.012066206,-1.450857877,2.208902209,1.364986853 2020 Q2,4.691348559,-8.52518022,-23.60556111,1.584258038 2020 Q3,4.203524981,-6.871101135,-13.99156399,1.80163077
,Food and beverages purchased for off-premises consumption,Clothing and footwear,Gasoline and other energy goods,Other nondurable goods 2019 Q1,0,0,0,0 2019 Q2,0.372376523,-5.209119689,31.70595588,1.588892278 2019 Q3,-0.168414569,3.348867078,-10.00888101,0.804191426 2019 Q4,0.768532598,-6.176747889,10.41566193,3.046165356 2020 Q1,3.106467805,2.619202001,-16.60900654,0.044961922 2020 Q2,15.3852134,-25.76736824,-68.79012086,0.868085535 2020 Q3,-1.850867504,7.431495609,60.66349869,0.858682015
,Nondurable Goods_Except Food,Foods_Feeds_and Beverages 2019 Q1,0,0 2019 Q2,3.428761651,-0.069783671 2019 Q3,3.695073236,2.651779484 2019 Q4,-1.165113182,2.30286113 2020 Q1,0.565912117,3.698534543 2020 Q2,-2.296937417,-0.976971389 2020 Q3,3.528628495,6.838799721
,Housing and utilities,Health care,Transportation services,Recreation services,Food services and accommodations,Financial services and insurance,Other services 2019 Q1,0,0,0,0,0,0,0 2019 Q2,0.338983051,0.969202899,1.149954002,0.940940941,1.116848805,-0.011712345,-0.072404743 2019 Q3,0.755840586,1.177536232,2.897884085,0.620620621,1.729314279,0.515343172,1.203728844 2019 Q4,0.737517178,2.364130435,2.989880405,1.441441441,1.717305152,1.405481377,2.31695176 2020 Q1,0.659642694,-2.09692029,-4.599816007,-8.368368368,-7.40963132,0.890138206,-0.959362838 2020 Q2,1.722400366,-19.26177536,-38.52345906,-50.75075075,-38.26107842,0.60904193,-18.75282831 2020 Q3,1.901053596,-5.108695652,-21.84912603,-33.87387387,-18.32592771,2.108222066,-12.44456512
,Wholesale trade,Retail trade,Transportation and warehousing,Information,Finance_insurance_real estate_rental_and leasing,Professional and business services,Educational services_health care_and social assistance,Arts_entertainment_recreation_accommodation_and food services,Other services_except government 2019 Q1,0,0,0,0,0,0,0,0,0 2019 Q2,-2.714567282,0.063068745,1.060445387,2.362400531,0.121369266,0.675811577,0.479747374,0.290939318,-0.160556596 2019 Q3,-2.831033865,1.126227588,2.757158006,4.127984085,0.041375886,1.926867533,1.050586021,1.898032696,0.481669789 2019 Q4,-2.956459416,1.810973962,3.30505479,5.968169761,0.89371914,2.711291685,1.633570171,1.468550845,0.053518865 2020 Q1,-3.010213223,-0.027029462,0.724637681,5.188992042,-0.45789314,2.224546442,-0.103236777,-5.971183153,-3.478726251 2020 Q2,-12.91883175,-8.937742139,-22.42841994,3.365384615,-1.095081786,-6.456414176,-14.72642254,-49.21030756,-22.63848006 2020 Q3,-4.389894284,0.25227498,-12.01838105,8.11505305,1.197142305,-0.993603926,-3.273213093,-26.71100028,-12.25582018
,Housing and utilities,Health care,Transportation services,Recreation services,Food services and accommodations,Financial services and insurance,Other services 2019 Q1,0,0,0,0,0,0,0 2019 Q2,0.778309885,0.4515087,0.76522606,0.493704179,0.809723486,1.499088092,0.083329518 2019 Q3,1.491278132,0.93421922,0.923071243,0.798664814,1.371086407,2.322771015,0.627260657 2019 Q4,2.224097003,1.525218413,1.725220152,1.876949695,1.585424977,2.725345116,0.862597866 2020 Q1,3.036318371,2.041883856,0.899994462,2.457161332,2.20292419,3.369611957,1.503594158 2020 Q2,3.592962954,2.999045591,-0.938763454,3.051353973,1.917989981,2.019543008,1.649191887 2020 Q3,4.085920118,3.786432714,-1.595067107,3.296021531,3.140230159,3.144229771,2.230667094
,Housing and utilities,Health care,Transportation services,Recreation services,Food services and accommodations,Financial services and insurance,Other services 2019 Q1,0,0,0,0,0,0,0 2019 Q2,3.149774472,1.818303265,3.096218077,1.989489541,3.278445861,6.132540863,0.333734932 2019 Q3,2.860020036,1.936062957,0.628059756,1.219386266,2.24609012,3.285798091,2.191699458 2019 Q4,2.919636642,2.362767342,3.217353736,4.348116684,0.848444388,1.583053858,0.938767772 2020 Q1,3.216278841,2.051205965,-3.205648186,2.297623299,2.45370782,2.5323964,2.566392974 2020 Q2,2.178539352,3.805157269,-7.092577939,2.340028204,-1.110515622,-5.122779912,0.574999536 2020 Q3,1.917068505,3.093085446,-2.623872628,0.95307931,4.883937975,4.48314901,2.307873451
,Domestic business,Households and Institutions,Government,Net lending or net borrowing (-),Gross domestic investment 2019 Q1,517.3,1354.5,-1229.5,-542.7,4503.5 2019 Q2,588.9,1186.1,-1208.2,-539.6,4489.6 2019 Q3,615,1180.6,-1303.9,-511.5,4517.7 2019 Q4,642.7,1203.8,-1277.3,-444.2,4505.4 2020 Q1,400,1595.3,-1353.5,-435.7,4464 2020 Q2,224.9,4770.8,-5168,-570.9,3914.9 2020 Q3,673.4,2834.7,-3594.2,-703.5,4482.7