1 Introduction
2 Conceptual framework
2.1 Disasters and entrepreneurial activity
2.2 Hypotheses
-
H1: Pandemic severity in a given region negatively affects new firm creation in that regionWe now turn attention to the effects of shutdown policies, which, as we mentioned above, mandate temporary shutdowns of specific nonessential economic activities to control the spread of the disease. We argue that such policies reduce new firms’ expected profits. In several industries (e.g. retail, restaurants, entertainment), shutdown policies result in a drop in revenue expectations because consumers cannot reach the places where products/services are typically sold or consumed (Baldwin & Weder di Mauro, 2020). The negative effects of shutdown policies on expected revenues are particularly severe for firms whose consumption is seasonal, as the sales lost during the shutdown are impossible to recoup after the shutdown has ended.2 Although consumers in some industries might change their behaviour and resort to online buying and consumption, the increase in online purchases is unlikely to compensate for the drop in physical sales and consumption, as data on transactions in the first months of the pandemic indicate (Andersen et al., 2020). In spite of this reduction in expected revenues, during shutdowns, firms in nonessential industries are still liable for their continuing fixed costs, such as rent, logistics and storage fees (Lu et al., 2020), which further erode expected profits. Moreover, when workers are not allowed to reach the workplace, expected labour productivity may be lower. Indeed, while some workers experience minor disruptions when switching to their home office, for many workers (e.g. those who require peculiar equipment to do their job), working from home means they cannot perform all their tasks (Adams-Prassl et al., 2020). Finally, in several industries, shutdown policies are expected to increase supply costs. For instance, during shutdowns, firms are likely to incur costly wastages of perishable production inputs if such inputs are not processed promptly after being procured.In line with these arguments, we claim that the effect of shutdown policies on new firm creation is negative, and we formulate the following hypothesis.
-
H2: Shutdown policies in a given industry and region negatively affect new firm creation in that industry and region.We expect this negative effect of shutdown policies on new firm creation to be greater in regions where the pandemic is more severe. As mentioned above, both pandemic severity and shutdown policies are expected to decrease new firms’ expected profits. In regions with greater pandemic severity, many prospective entrepreneurs have negative expectations about future profits; hence, they will refrain from creating new firms also in the absence of shutdown policies. Therefore, the additional reduction of expected profits engendered by the implementation of shutdown policies has negligible effects on new firm creation in these regions. Conversely, in regions where pandemic severity is lower, shutdown policies may hinder new firm creation by many prospective entrepreneurs who would not refrain from creating new firms in the absence of such policies. Following these arguments, we conclude that the negative effects of shutdown policies predicted by H2 are less evident in regions where pandemic severity is greater than in the regions less affected by the pandemic.We now propose an additional argument in favour of the reduced negative effect of shutdown policies in regions with greater pandemic severity. As shutdown policies are implemented to limit the spread of the disease, prospective entrepreneurs in these regions might expect that the temporary closure of economic activities will help in containing the contagion, thus reducing pandemic severity in the near future and significantly improving the disaster situation. Such expectations of noticeable future reductions in pandemic severity may positively affect new firms’ expected profits and, to some extent, counterbalance the negative effects of shutdown policies on expected profits. Conversely, in regions where pandemic severity is lower, shutdown policies are probably expected to result in less significant improvements in the disaster situation. These expectations of future unremarkable reductions in pandemic severity after implementing shutdown policies may have negligible positive effects on new firms’ expected profits that are thus unable to compensate for the negative effects of the shutdown. Following these arguments, we claim that the negative effects of shutdown policies on firm creation will be weaker in the regions more affected by the pandemic than in the less affected regions. Hypothesis 3 follows.
-
H3: The negative relation between shutdown policies and new firm creation in a given industry and region is weaker the greater the pandemic severity in the region.Let us now discuss the effects on new firm creation of policy measures aimed at protecting the economic system by funding specific industries hit by the pandemic. These policies vary depending on their direct beneficiaries, which may be either consumers or existing firms. We first consider demand stimulus policies having consumers as direct beneficiaries. These policies are aimed at stimulating the demand for specific nonessential products or services, e.g. by providing consumers with vouchers to purchase these products/services or granting them tax credits after purchase. Demand stimulus policies may positively affect the expectations about new firms’ short-term profits. As tax rebates and, in particular, shopping vouchers tend to have some positive effects on consumers’ spending (e.g. Hsieh et al., 2010; Kan et al., 2017), prospective entrepreneurs may anticipate that the demand in the industries supported by demand stimulus policies will increase in the near future. Hence, prospective entrepreneurs will be more likely to create new firms in these industries. Based on these arguments, we formulate the following hypothesis.
-
H4: Demand stimulus policies in a given industry positively affect new firm creation in that industryThe positive effects on new firm creation of demand stimulus policies are likely to depend on pandemic severity. In regions with greater pandemic severity, individuals experience stronger emotions of fear and anxiety (Le & Nguyen, 2021). Such negative emotions create irregular and often irrational consumer behaviours (Loxton et al., 2020). Prospective entrepreneurs may expect that consumers who do not have entirely rational behaviours may be less responsive than rational consumers to the stimuli of policies aimed at increasing demand for nonessential products/services; hence, demand stimulus policies are unlikely to significantly increase consumers’ spending. Conversely, in regions where pandemic severity is lower, individuals experience less intense negative emotions that are less likely to generate irrational consumer behaviours. Hence, prospective entrepreneurs might expect that policy stimuli will have stronger effects in these regions. Following these arguments, we claim that pandemic severity reduces the positive effect of demand stimulus policies on new firm creation. Hypothesis 5 follows.
-
H5: The positive relation between demand stimulus policies in a given industry and new firm creation in that industry is weaker the greater the pandemic severity in the regionOther policy measures aimed at protecting the economic system are those having existing firms in specific industries as direct beneficiaries. Firm support policies encompass measures such as grants to cover their expenses or to make investments, relief funds aimed at compensating for missed revenues, tax exemptions and temporary suspensions of fee payments. As firm support policies target only the firms that were active at the outbreak of the pandemic, they likely have negative effects on new firm creation. On the one hand, firm support policies engender cost advantages for existing firms with respect to prospective entrants that cannot benefit from these measures. As cost advantages of existing firms serve as important barriers to entry (e.g. Bain, 1956; Porter, 1980), prospective entrepreneurs may be discouraged from entering the industries targeted by firm support policies. On the other hand, firm support policies probably reduce existing firms’ failure rates, thus reducing the amount of resources freed by exited firms and, as a consequence, the probability that new firms are created in the industry to take advantage of these resources. These arguments lead to the following hypothesis.
-
H6: Firm support policies in a given industry negatively affect new firm creation in that industryPandemic severity may also affect the negative relationship between new firm creation and firm support policies in given industries, even though this moderating effect is hard to predict. On the one hand, we expect that the negative effect of firm support policies on the creation of new firms will be weaker the greater the pandemic severity. In line with the arguments leading to H1 and H3, in the regions with greater pandemic severity, many prospective entrepreneurs have negative expectations about future profits because of the pandemic. Thus, they will refrain from creating new firms, regardless of firm support policies. Conversely, in regions that are less hit by the pandemic, providing existing firms with cost advantages is likely to reduce the number of prospective entrepreneurs willing to create new firms. On the other hand, we might claim that the negative effect of firm support policies on new firm creation is more intense as pandemic severity increases. In the regions with greater pandemic severity, the probability of failure of existing firms is very high (Amankwah-Amoah et al., 2021); hence, firm support policies may contribute to reducing existing firms’ allegedly high failure rates. Conversely, firm support policies are less likely to reduce existing firms’ failure rates in the least severely affected regions, where existing firms are probably less likely to fail.As opposing forces are at work, we do not formulate any hypothesis on the moderating effect of pandemic severity on the relationship between new firm creation and firm support policies.Our six hypotheses are synthesized in Fig. 1.
3 The Italian context during the COVID-19 pandemic
3.1 The spread of COVID-19 in Italy and the policies taken to control the spread of the disease
3.2 The policies approved by the Italian government to protect the economy
4 Data and descriptive evidence on the Italian case
2019–2020 variation | COVID-19 deaths per 10,000 inhabitants | |
---|---|---|
North | ||
Aosta Valley | − 6.0% | 30.20 |
Emilia-Romagna | − 15.0% | 17.32 |
Friuli-Venezia Giulia | − 16.7% | 13.56 |
Liguria | − 18.3% | 18.73 |
Lombardy | − 13.4% | 24.86 |
Piedmont | − 14.1% | 18.25 |
Trentino-South Tyrol | − 13.0% | 15.64 |
Veneto | − 11.8% | 13.32 |
Centre | ||
Latium | − 19.4% | 6.43 |
Marche | − 18.1% | 10.35 |
Tuscany | − 17.1% | 9.87 |
Umbria | − 15.8% | 7.09 |
South and islands | ||
Abruzzo | − 16.4% | 9.29 |
Apulia | − 15.7% | 6.17 |
Basilicata | − 18.6% | 4.60 |
Calabria | − 14.8% | 2.45 |
Campania | − 13.4% | 4.92 |
Molise | − 16.7% | 6.32 |
Sardinia | − 16.5% | 4.58 |
Sicily | − 16.0% | 4.85 |
Industry (NACE code) | 2019–2020 variation |
---|---|
Sports activities and amusement and recreation activities (R 93) | − 39.49% |
Food and beverage service activities (I 56) | − 36.43% |
Manufacture of fabricated metal products, except machinery and equipment (C 25) | − 28.00% |
Wholesale and retail trade and repair of motor vehicles and motorcycles (G 45) | − 23.43% |
Other personal service activities (S 96) | − 23.23% |
Accommodation (I 55) | − 23.14% |
Manufacture of machinery and equipment (C 28) | − 21.40% |
Warehousing and support activities for transportation (H 52) | − 19.11% |
Electricity, gas, steam and air conditioning supply (D 35) | − 18.35% |
Repair and installation of machinery and equipment (C 33) | − 14.71% |
Office administrative, office support and other business support activities (N 82) | − 13.81% |
Agriculture, forestry and fishing (A 01) | − 13.35% |
Manufacture of wearing apparel (C 14) | − 9.24% |
Building completion and finishing (F 43) | − 8.60% |
Real estate activities (L 68) | − 6.76% |
Wholesale trade, except of motor vehicles and motorcycles (G 46) | − 6.44% |
Rental and leasing activities (N 77) | − 6.29% |
Manufacture of food products (C 10) | − 6.22% |
Retail trade, except of motor vehicles and motorcycles (G 47) | − 5.35% |
Advertising and market research (M 73) | − 3.79% |
Scientific research and development (M 72) | − 1.83% |
Activities of head offices (M 70) | 0.72% |
Other professional, scientific and technical activities (M 74) | 1.06% |
Construction of buildings (F 41) | 2.41% |
Computer programming, consultancy and related activities (J 62) | 7.53% |
Services to buildings and landscape activities (N 81) | 8.14% |
Information service activities (J 63) | 8.26% |
Financial service activities, except insurance and pension funding (K 64) | 9.62% |
Architectural and engineering activities; technical testing and analysis (M 71) | 11.46% |
Education (P 85) | 15.23% |
5 Econometric models
Variable | Mean | SD | (1) | (2) | (3) | (4) | (5) | (6) | |
---|---|---|---|---|---|---|---|---|---|
(1) | NewFirmCreationDensityijt | 0.05 | 0.14 | 1.00 | |||||
(2) | COVID19Severityit | 0.10 | 0.17 | − 0.04 | 1.00 | ||||
(3) | ShutdownPoliciesijt | 0.08 | 0.24 | − 0.06 | 0.21 | 1.00 | |||
(4) | DemandStimulusPoliciesjt | 0.08 | 0.38 | 0.09 | 0.00 | − 0.02 | 1.00 | ||
(5) | FirmSupportPoliciesit | 0.05 | 0.29 | 0.22 | 0.02 | 0.03 | 0.14 | 1.00 | |
(6) | PriorFirmCreationDensityijt | 0.05 | 0.14 | 0.70 | − 0.03 | 0.04 | 0.22 | 0.08 | 1.00 |
(7) | TestsDensityit | 0.04 | 0.04 | 0.00 | 0.53 | − 0.06 | 0.07 | 0.08 | − 0.05 |
6 Results
6.1 Main econometric estimates
Model 1 January–December period | Model 2 January–August period | Model 3 September–December period | |
---|---|---|---|
Constant | 0.019 (0.003) *** | 0.033 (0.004) *** | 0.006 (0.004)* |
PriorFirmCreationDensityijt | 0.477 (0.036)*** | 0.455 (0.045) *** | 0.645 (0.059)*** |
TestsDensityit | 0.183 (0.032)*** | − 0.371 (0.085) *** | 0.224 (0.080)*** |
COVID19Severityit | − 0.050 (0.009)*** | − 0.088 (0.010)*** | − 0.019 (0.017) |
No. of observations | 20,412 | 13,608 | 6,804 |
No. of region-industry pairs | 1,701 | 1,701 | 1,701 |
No. of months | 12 | 8 | 4 |
Overall R2 | 0.496 | 0.495 | 0.520 |
Model 1 January–December period | Model 2 January–August period | Model 3 September–December period | |
---|---|---|---|
Constant | 0.025 (0.003)*** | 0.036 (0.004)*** | 0.006 (0.006) |
PriorFirmCreationDensityijt | 0.471 (0.038)*** | 0.443 (0.048)*** | 0.645 (0.060)*** |
TestsDensityit | 0.096 (0.030)*** | − 0.380 (0.077)*** | 0.221 (0.082)*** |
COVID19Severityit | − 0.022 (0.009)** | − 0.039 (0.010)*** | − 0.017 (0.016) |
ShutdownPoliciesijt | − 0.049 (0.005)*** | − 0.051 (0.005)*** | − 0.004 (0.013) |
DemandStimulusPoliciesjt | 0.009 (0.004)** | 0.014 (0.005)*** | 0.028 (0.032) |
FirmSupportPoliciesit | − 0.022 (0.015) | − 0.032 (0.019)* | − 0.039 (0.025) |
No. of observations | 20,412 | 13,608 | 6,804 |
No. of region-industry pairs | 1,701 | 1,701 | 1,701 |
No. of months | 12 | 8 | 4 |
Overall R2 | 0.488 | 0.484 | 0.480 |
Model 1 January–December period | Model 2 January–August period | Model 3 September–December period | |
---|---|---|---|
Constant | 0.024 (0.003)*** | 0.036 (0.004)*** | 0.001 (0.008) |
PriorFirmCreationDensityijt | 0.470 (0.038)*** | 0.446 (0.048)*** | 0.635 (0.059)*** |
TestsDensityit | 0.109 (0.030)*** | − 0.383 (0.077)*** | 0.240 (0.078)*** |
COVID19Severityit | − 0.023 (0.009)*** | − 0.048 (0.010)*** | − 0.013 (0.013) |
ShutdownPoliciesijt | − 0.058 (0.007)*** | − 0.059 (0.006)*** | − 0.038 (0.044) |
DemandStimulusPoliciesjt | 0.009 (0.005)** | 0.019 (0.005)*** | 0.034 (0.034) |
FirmSupportPoliciesit | − 0.013 (0.015) | − 0.015 (0.016) | 0.015 (0.040) |
ShutdownPoliciesijt × COVID19Severityit | 0.050 (0.029)* | 0.071 (0.024)*** | 0.121 (0.156) |
DemandStimulusPoliciesjt × COVID19Severityit | − 0.003 (0.026) | − 0.196 (0.063)*** | − 0.047 (0.047) |
FirmSupportPoliciesit × COVID19Severityit | − 0.082 (0.030)*** | − 0.376 (0.106)*** | − 0.072 (0.039)* |
No. of observations | 20,412 | 13,608 | 6804 |
No. of region-industry pairs | 1701 | 1701 | 1701 |
No. of months | 12 | 8 | 4 |
Overall R2 | 0.489 | 0.484 | 0.504 |