Chapter 9. Employment and productivity dynamics during economic crises in Japan

Kenta Ikeuchi

Research Institute of Economy, Trade and Industry (RIETI)

This chapter examines the effects of economic crises on employment dynamics in Japan, in which during the last two decades, the economy has been going through a long stagnation and suffered a number of economic crises. Focusing on four crises during the period and utilising a comprehensive panel dataset of Japanese listed companies, this chapter considers the effects of these crises on the firm-level within-industry reallocation effects. The results show that the reallocation of labour inputs was productivity-enhancing in Japan and the economic crises reinforced the productivity-enhancing reallocation mechanisms, in both the manufacturing and non-manufacturing sectors. However, the global financial crisis at the end of 2000s did not strengthen these mechanisms. These results are consistent with existing empirical findings in the United States.

Introduction

Job creation has been one of the most important and pressing issues in governments’ policy agendas across the OECD. In order to meet this policy interest, a deep understanding of employment dynamics is a critical research issue. Since the early 1990s, Japan has experienced a slowdown in productivity and economic growth and thus job creation. Reallocation of resources across firms is a key mechanism for productivity and economic growth. Reallocation of production factors, such as labour inputs, from a relatively low productive firm to a highly productive firm increases productivity in that sector and at the macro level, which thus impacts on job creation. Understanding such interrelationships between the dynamics of employment and productivity is one of the key objectives of the OECD DynEmp and MultiProd projects.

This raises the issue of the impact of crises. Do economic crises have market cleansing effects? This is a long standing and still ongoing debate (Foster, Grim and Haltiwanger, 2016). According to the cleansing hypothesis, recessions reinforce productivity-enhancing reallocation through their associated low adjustment costs (Davis and Haltiwanger, 1990; Caballero and Hammour, 1994; Mortensen and Pissarides, 1994). There are also alternative hypotheses related to recessions that highlight their potential distortions of reallocation dynamics (Caballero and Hammour, 1996) and “sullying” or “scarring” effects (Osotimehin and Pappadà, 2016). Barlevy (2003) argues that the cleansing effect can be reversed also when financial constraints are present.

There are various empirical studies on the relationship between reallocation effects and economic crises. Some studies obtain results consistent with the cleansing hypothesis (Davis and Haltiwanger, 1992, 1999; Davis, Faberman and Haltiwanger, 2006, 2012).

Investigating the link between credit booms, productivity growth, labour reallocations, and financial crises in a sample of over 20 advanced economies and over 40 years, Borio et al. (2015) found that i) credit booms tend to undermine productivity growth by inducing labour reallocations towards lower productivity growth sectors; and ii) the impact of reallocations that occur during a boom, and during economic expansions more generally, is much larger if a crisis follows.

Using establishment-level micro-data for the United States, Foster, Grim, and Haltiwanger (2016) have found that downturns prior to the global financial crisis (GFC) are periods of accelerated reallocation that are even more productivity-enhancing than reallocation in normal times, but during the recent (2007-09) GFC, the intensity of reallocation fell and the reallocation that did occur was less productivity-enhancing than in prior recessions.

Lucchese and Pianta (2012) using cross-country data show that during downturns implementing new processes contributes to restructuring and job losses.

The collapse of Japan’s bubble economy in the early 1990s was followed by a long economic stagnation and by financial crises. During this period of recession, many Japanese banks continued to lend to otherwise insolvent firms (Caballero, Hoshi and Kashyap, 2008). Such “zombie lending” is expected to reduce the productivity-enhancing reallocation effects, since resources are stuck in unprofitable and unproductive firms. Kwon, Narita and Narita (2015) investigate the amount of aggregate output growth that was driven by resource reallocation and how much more would have been generated had there been no zombie lending during the 1990s. They found that the contribution to aggregate productivity growth of resource reallocation deteriorated in the 1990s and became negative during the late 1990s, when the Asian financial crisis occurred. Using the industry-level EU KLEMS database, Fukao, Miyagawa and Takizawa (2007) and Fukao et al. (2009) looked at the sources of economic growth for Japan and the Republic of Korea during the period 1975-2005, and they found that the resource reallocation effects of capital input in both Japan and Korea were either negligible or insignificant, while those of labour input (the labour shift from lower wage industries to higher wage industries) were positive and significant. They concluded that a series of productivity-enhancing policies designed to promote the reallocation of capital input seems crucial to resume sustainable growth paths.

Fukao, Kim, and Kwon (2008) analysed the total factor productivity (TFP) growth rate of the Japanese manufacturing sector from 1981-2003 and found that the reallocation of resources from less efficient to more efficient firms was very slow and limited. They emphasised that the “low metabolism” seems to be an important cause for the slowdown in Japan’s TFP growth. Using original Japanese enterprise-level data for the Financial Statements Statistics of Corporations by Industry for 1982-2007, Inui et al. (2011) observed TFP trends in both the manufacturing and non-manufacturing industries and found that the acceleration of the TFP growth rate is mainly determined by TFP growth within firms, and that the contribution of resource reallocations across firms to aggregated TFP growth are comparably small, especially in the manufacturing sector. Recently, Hosono and Takizawa (2015) found that there are distortions in the reallocation of production factors in Japanese manufacturing industries and that such distortions have a significant impact on entry and exit as well as on establishment-level productivity growth. They also found that financial constraints play a significant role as factors of distortion.

With these remarks as background and motivation, this chapter addresses empirical questions concerning the potential cleansing effects of the recent economic crises in Japan. First, is reallocation of labour inputs enhancing productivity? Second, is the relationship between productivity and reallocation influenced by economic crises? Third, is the relationship between productivity and reallocation different across crises?

This chapter is organised as follows. The next section overviews the economic crises and their impacts on the labour market in Japan from the 1990s to 2010s. Then, the fourth section examines the interrelationship between employment dynamics and productivity dynamics in Japan during some of the economic crises that have occurred during the last three decades or so (from 1980). The last section concludes the chapter by proposing a future research agenda related to a global perspective on employment dynamics in Japan.

Economic crises in Japan over three decades

This section overviews the several macro events that have potentially affected the employment dynamics during the last three decades in Japan. It focusses particularly on the effects of four economic crises: the burst of the bubble economy in the early 1990s, the Asian financial crisis in the late 1990s, the information technology (IT) bubble burst in the early 2000s, and the GFC at the end of the 2000s.

Figure 9.1 shows the trend of total real value-added to the market economy in Japan in the period 1980 to 2012. Surprisingly, the figure clearly indicates that Japanese real GDP only increased by 8% in the 21 years from 1991-2012, while in the 1980s it had grown by 60%. During the whole period, the Japanese economy had experienced negative real value-added growth five times. First, the burst of the bubble economy occurred in 1992, and the real GDP decreased until 1994. Although from 1995 it recovered, and the economy began to grow again, in 1998 the Asian financial crisis caused damage to the Japanese economy. From 1997-99, the real value-added decreased again. Soon after its recovery in 2000, the IT bubble burst occurred, and the Japanese economy slowed down between 2000 and 2001. The GFC in 2007-09 has had the largest impact, with total real value-added decreasing by more than 10% over the period 2007-09.

After the bubble burst in 1992, the labour demand had been constantly decreasing in Japan. While the labour population did not decrease from 1991-2012, actual labour input (man hours) declined by more than 20%. It seems that the crises accelerate the declining trend of labour demand. During the five periods of crisis (1991-94, 1997-99, 2000-01 and 2007-09), man hours declined more than it did outside of these periods. From the viewpoint of the supply side, the working population also began to decrease from 1997, reflecting the declining birth rate and ageing population.

TFP in Japan has also been stagnant during the last two decades. In contrast to gross domestic product (GDP) and labour demand, however, TFP did not decline during the crisis following the IT bubble burst (2000-01), while during the other three crises TFP declined, along with the economic contraction.

Figure 9.2 presents the effects of crises on labour demand by job status. This shows that labour demand for the self-employed has constantly decreased. During the Japanese financial crisis (1991-94), demand for regular workers and part-time workers increased. While the working hours of part-time workers also increased during the Asian financial crisis and the IT bubble burst, demand for regular workers declined. The GFC, however, affected part-time jobs more than regular workers.

In Figure 9.3, gross-output growth rates during the crises are decomposed into several final demand factors. During the first crisis (Japanese financial crisis), public expenditure increased. The GFC can be characterised by a huge decrease in net exports and household expenditures in addition to the decrease in private investments.

Firm-level reallocation and crises

This section relates employment dynamics to productivity for a subset of the population of Japanese firms, in particular it limits its analysis to listed companies in Japan. Utilising the micro dataset of listed firms1 it investigates whether the macroeconomic crises have enhanced within-industry reallocation effects. Figure 9.4 shows how the number of listed companies, that is, the size of this study’s sample, has evolved over time. The number of listed firms increased from 1980 to 2006. There are between 1 600 and 3 856 firms for each year. Figure 9.5 shows the coverage ratio of the listed company sample in the total economy. The listed companies cover almost 7% to 9% of total employees and 15% to 18% of total value-added in Japan.

For the period 1980-2012, there were 4 799 active firms in the sample (Table 9.1). The firms in the sample vary across industries: 2 115 firms were in manufacturing, and 2 706 firms were in non-manufacturing industries (Table 9.2). The share of the number of non-manufacturing2 firms increased over time.

Table 9.1. Number of firms: manufacturing industries
Industry	1980-2012	80-89	90-99	00-12
Total	4 849	2 308	3 488	4 606
Manufacturing	2 143	1 418	1 812	2 013
Livestock products	28	21	25	28
Seafood products	8	2	8	8
Flour and grain mill products	7	7	7	7
Miscellaneous foods and related products	131	77	109	127
Prepared animal foods and organic fertilisers	7	5	7	7
Beverages	9	9	9	8
Textile products	81	64	76	75
Lumber and wood products	13	7	9	12
Furniture and fixtures	14	9	11	13
Pulp, paper, and coated and glazed paper	30	24	25	23
Paper products	22	13	20	20
Printing, plate making for printing and bookbinding	29	13	26	28
Leather and leather products	4	3	3	4
Rubber products	25	22	24	24
Chemical fertilisers	6	5	5	5
Basic inorganic chemicals	42	35	40	38
Basic organic chemicals	6	5	5	5
Organic chemicals	59	45	53	57
Chemical fibres	7	7	7	7
Miscellaneous chemical products	90	53	76	89
Pharmaceutical products	88	48	56	85
Petroleum products	14	10	10	13
Coal products	2	2	1	1
Glass and its products	16	12	13	13
Cement and its products	39	26	36	31
Pottery	13	12	13	13
Miscellaneous ceramic, stone and clay products	37	29	33	32
Pig iron and crude steel	46	37	36	40
Miscellaneous iron and steel	32	27	31	29
Smelting and refining of non-ferrous metals	22	13	15	21
Non-ferrous metal products	43	34	36	37
Fabricated constructional and architectural metal products	49	30	44	46
Miscellaneous fabricated metal products	71	46	67	68
General industry machinery	79	66	70	75
Special industry machinery	132	92	110	123
Miscellaneous machinery	79	56	68	72
Office and service industry machines	25	16	22	22
Electrical generating, transmission, distribution and industrial apparatus	54	40	50	53
Household electric appliances	42	28	35	38
Electronic data processing machines, digital and analogue computer equipment and accessories	25	9	17	24
Communication equipment	38	32	35	37
Electronic equipment and electric measuring instruments	42	22	34	40
Semiconductor devices and integrated circuits	24	10	16	22
Miscellaneous electrical machinery equipment	132	72	101	128
Motor vehicles	15	13	14	15
Motor vehicle parts and accessories	121	85	112	117
Other transportation equipment	28	24	24	25
Precision machinery and equipment	79	44	63	72
Plastic products	58	28	51	58
Miscellaneous manufacturing industries	80	29	54	78
Source: Source: Author’s calculation based on Research Institute of Capital Formation, Development Bank of Japan (2015), Corporate Financial Data Bank (database), www.dbj.jp/ricf/databank/.

Table 9.2. Number of firms: non-manufacturing industries
Industry	1980-2012	80-89	90-99	00-12
Total	4 849	2 308	3 488	4 606
Non-manufacturing	2 706	890	1 676	2 593
Fisheries	1	1	1	1
Mining	13	7	8	13
Construction	289	169	256	271
Civil engineering	5	5	5	4
Electricity	13	10	10	13
Gas, heat supply	19	13	16	19
Waste disposal	5		1	5
Wholesale	474	193	367	443
Retail	434	152	305	411
Finance	5	2	2	5
Insurance	1			1
Real estate	246	62	90	235
Railway	27	24	27	27
Road transportation	59	28	49	59
Water transportation	46	41	43	36
Air transportation	9	5	6	9
Other transportation and packing	38	20	31	37
Telegraph and telephone	36	2	18	35
Education (private and non-profit)	30	3	21	29
Research (private)	1			1
Medical (private)	9	1	2	9
Hygiene (private and non-profit)	1		1	1
Advertising	45	1	10	44
Rental of office equipment and goods	24	8	19	24
Automobile maintenance services	1		1	1
Other services for businesses	92	8	39	92
Entertainment	37	19	24	36
Broadcasting	21	6	11	21
Information services and internet-based services	476	49	180	470
Publishing	14	5	11	13
Video picture, sound information, character information production and distribution	27	11	14	25
Eating and drinking places	134	26	74	132
Accommodation	18	15	17	15
Laundry, beauty and bath services	9	2	4	9
Other services for individuals	29	1	10	29
Social insurance and social welfare (non-profit)	18	1	3	18
Source: Source: Author’s calculation based on Research Institute of Capital Formation, Development Bank of Japan (2015), Corporate Financial Data Bank (database), www.dbj.jp/ricf/databank/.

Following Fukao et al. (2011), the cross-sectional TFP index for each firm is calculated as the relative value of the industry average TFP in each year. Figure 9.6 compares the employment growth rates between firms that had a relatively high TFP with those that had a lower TFP. In almost the whole period, the average employment growth rate of the firms in the highest quartile for TFP (upper 25%) was higher than that those in the lowest quartile (lower 25%). During the periods of crisis, the difference in the average employment growth rates between highest and lowest quartiles (difference between the upper 25% and the lower 25%) tended to increase.

In order to examine the effects of economic crises on reallocation effects, this analysis relates employment growth rate of firms to their TFP. Table 9.3 shows the results of a regression analysis in which the dependent variable is the firm-level employment growth rate for the sample companies during 1980-2012. Column [1] of the table shows the basic result with relative employment size and TFP as independent variables. The independent variables in all the models include a set of year-industry dummies, in order to factor out the effects of time varying industry-specific market conditions. The coefficient for relative employment size is negative and statistically significant, and the coefficient for relative TFP is positive and significant. The results indicate that the employment growth rate of large firms tends to be lower than that of smaller firms. The positive coefficient for relative TFP indicates that reallocation is productivity-enhancing in general. For a given firm size, firms with a higher TFP grow faster than those with a lower one.

Table 9.3. Reallocation effects and economic crisis
Dependent variable: employment growth rate
	[1]	[2]	[3]	[4]	[5]
Ln. emp. size	-0.004***	-0.004***	-0.004***	-0.004***	-0.004***
	[0.000]	[0.000]	[0.000]	[0.001]	[0.001]
Ln. TFP	0.148***	0.149***	0.149***	0.142***	0.144***
	[0.005]	[0.005]	[0.005]	[0.006]	[0.006]
*Ln. emp. size industry growth**		-0.006	0	-0.004	0
		[0.004]	[0.005]	[0.005]	[0.005]
*Ln. TFP industry growth**		-0.107**	-0.108*	-0.059	-0.113**
		[0.050]	[0.062]	[0.054]	[0.057]
*Ln. emp. size GDP growth**			-0.020**
			[0.010]
*Ln. TFP GDP growth**			0.009
			[0.115]
*Ln. emp. size crisis dummy**				0.001
				[0.001]
*Ln. TFP crisis dummy**				0.022**
				[0.009]
*Ln. emp. size Japanese financial crisis (91-94)**					0.003***
					[0.001]
*Ln. emp. size Asian financial crisis (97-99)**					-0.004***
					[0.001]
*Ln. emp. size IT bubble burst (00-01)**					-0.004**
					[0.002]
*Ln. emp. size GFC (07-09)**					0.005***
					[0.001]
*Ln. TFP Japanese financial crisis (91-94)**					0.025*
					[0.013]
*Ln. TFP Asian financial crisis (97-99)**					0.068***
					[0.016]
*Ln. TFP IT bubble burst (00-01)**					0.044**
					[0.019]
*Ln. TFP GFC (07-09)**					-0.024
					[0.015]
Constant	-0.131***	-0.134***	-0.127***	-0.132***	-0.129***
	[0.003]	[0.004]	[0.005]	[0.004]	[0.005]
Industry-year dummies	Yes	Yes	Yes	Yes	Yes
Industry dummies	Yes	Yes	Yes	Yes	Yes
Year dummies	Yes	Yes	Yes	Yes	Yes
No. of observations	79 000	79 000	79 000	79 000	79 000
No. of firms	4 551	4 551	4 551	4 551	4 551
R2	0.163	0.163	0.163	0.163	0.164
Notes: Notes: * p < 0.01, p < 0.05, * p < 0.1. Highest and lowest 1% outliers are removed.
Source: Source: Author’s calculation based on Research Institute of Capital Formation, Development Bank of Japan (2015), Corporate Financial Data Bank (database), www.dbj.jp/ricf/databank/and Research Institute of Economy, Trade and Industry (2015), The Japan Industrial Productivity Database 2015 (database), www.rieti.go.jp/en/about/about.html.

Next, this study examines whether such a market cleansing mechanism becomes stronger during an economic crisis. Column [2] of Table 9.3 includes the interaction terms of industry (gross-output) growth rate with employment size and TFP, in addition to the main effects of these variables.3 The coefficient for the interaction term of TFP and industry growth is significantly negative. This result indicates that an industry downturn reinforces the productivity-enhancing effects of reallocation. In contrast, the coefficient for the interaction term of TFP and GDP growth rate is not significant (column [3]). These results imply that a macro level economic downturn has no negative impact on inter-firm reallocation effects in industries which do not face a demand reduction while a demand shrink in a particular industry reinforces productivity-enhancing reallocation in that industry.

Column [4] examines the effects of economic crises utilising a dummy variable that takes the value of one in each of the four economic crisis periods in Japan, and, is zero otherwise. Here, a positive and statistically significant coefficient for the interaction term of TFP and the crisis dummy can be observed. This result indicates that an economy-wide crisis enhances the reallocation mechanism. Moreover, in column [5], the effects of each crisis are distinguished. Significantly positive coefficients are visible for the interaction terms of TFP and the dummy variables for the three crises (Japanese financial crisis, Asian financial crisis, and IT bubble burst), while the interaction term with the GFC dummy is not significant. This indicates that the effect of the GFC on the reallocation mechanism differs from earlier crises.

The effects of the economic crises on the reallocation mechanism may differ across industries. In order to check this possibility, Table 9.4 shows the estimation results of the same regression models as Table 9.3 but with firms in the sample divided into two subsamples according to industry. The results for manufacturing firms are shown in columns 1 and 2, and the results for non-manufacturing firms are shown in columns 3 and 4. The coefficient for TFP is positive and significant for both samples. This implies that the reallocation mechanism is productivity-enhancing in both manufacturing and non-manufacturing sectors. There are, however, several differences between the results for manufacturing and non-manufacturing sectors.

Table 9.4. Reallocation effects and economic crisis by sector
Dependent variable: employment growth rate
	Manufacturing		Non-manufacturing
	[1]	[2]	[3]	[4]
Ln. emp. size	-0.003***	-0.003***	-0.005***	-0.005***
	[0.000]	[0.001]	[0.001]	[0.001]
Ln. TFP	0.212***	0.196***	0.124***	0.122***
	[0.008]	[0.009]	[0.007]	[0.008]
*Ln. emp. size industry growth**	0	0.003	-0.024**	-0.013
	[0.004]	[0.005]	[0.010]	[0.011]
*Ln. TFP industry growth**	-0.04	0.014	-0.167**	-0.196**
	[0.063]	[0.078]	[0.081]	[0.092]
*Ln. emp. size Japanese financial crisis (91-94)**		0.005***		0.001
		[0.001]		[0.002]
*Ln. emp. size Asian financial crisis (97-99)**		-0.006***		-0.002
		[0.002]		[0.002]
*Ln. emp. size IT bubble burst (00-01)**		-0.006***		-0.003
		[0.002]		[0.003]
*Ln. emp. size GFC (07-09)**		0.004**		0.005***
		[0.002]		[0.002]
*Ln. TFP Japanese financial crisis (91-94)**		0.073***		0.003
		[0.023]		[0.015]
*Ln. TFP Asian financial crisis (97-99)**		0.155***		0.039**
		[0.027]		[0.020]
*Ln. TFP IT bubble burst (00-01)**		0.094***		0.031
		[0.031]		[0.023]
*Ln. TFP GFC (07-09)**		-0.029		-0.018
		[0.028]		[0.018]
Constant	-0.005	-0.005	0.032	0.031
	[0.034]	[0.034]	[0.064]	[0.064]
Industry-year dummies	Yes	Yes	Yes	Yes
Industry dummies	Yes	Yes	Yes	Yes
Year dummies	Yes	Yes	Yes	Yes
No. of observations	45 000	45 000	35 000	35 000
No. of firms	2 079	2 079	2 472	2 472
R2	0.19	0.192	0.141	0.142
Note: Note: * p < 0.01, p < 0.05, * p < 0.1. Highest and lowest 1% outliers are removed.
Source: Source: Author’s calculation based on Research Institute of Capital Formation, Development Bank of Japan (2015), Corporate Financial Data Bank (database), www.dbj.jp/ricf/databank/and Research Institute of Economy, Trade and Industry (2015), The Japan Industrial Productivity Database 2015 (database), www.rieti.go.jp/en/about/about.html.

First, the interaction term of TFP and industry gross-output growth rate is insignificant for manufacturing industries but significant and negative for non-manufacturing. Second, the Japanese financial crisis and IT bubble burst reinforced the productivity-enhancing reallocation mechanism only for the manufacturing sector. In contrast, both in manufacturing and non-manufacturing sectors, the Asian financial crisis reinforced the reallocation mechanism, while the GFC had no significant effect on this.

Conclusion

This chapter has examined the effects of economic crises on employment dynamics in Japan. During the last two decades (from 1990), the Japanese economy has been going through a long stagnation and suffered a number of economic crises. Focusing on four crisis periods, in which a negative growth rate of value-added was observed in the total market economy, this study has considered the effects of these on the labour market and on productivity. During the crises, both the labour inputs and TFP decreased sharply. Even in the economic recovery periods following the crises, when TFP increased, the labour inputs did not increase. In particular, this study found that demand for self-employed and regular workers was diminished by the crises, while the demand for part-time workers increased.

Then, utilising a comprehensive panel dataset of Japanese listed companies, it examined firm-level within-industry reallocation effects. During the period from 1980-2012, it found that the reallocation of labour inputs was productivity-enhancing in Japan. The results of regression analyses based on the firm-level panel data show the economic crises as having reinforced the productivity-enhancing reallocation mechanisms, in both the manufacturing and non-manufacturing sectors. However, it found that during the GFC at the end of 2000s, the productivity-enhancing reallocation mechanism was not strengthened. These results are consistent with existing empirical findings in the United States (Foster, Grim, and Haltiwanger, 2016).

The GFC caused a fluctuating global financial market and brought a sharp decline of net exports from the Japanese economy. Since highly productive firms tend to be more internationalised, they might also be more affected by such a downturn in the global economy. The results of this chapter may indicate that the market cleansing effects of an economic crisis in the era of high globalisation largely depend on international market conditions, rather than those of the domestic economy. To further investigate such a mechanism, a comparably rich international dataset is needed that is, for employment and productivity dynamics, linked to global value-chain data.

References

Barlevy, G. (2003), “Credit market frictions and the allocation of resources over the business cycle”, Journal of Monetary Economics, Vol. 50/8, pp. 1795-1818, https://doi.org/10.1016/j.jmoneco.2002.11.001.

Borio, C. et al. (2015), “Labour reallocation and productivity dynamics: financial causes, real consequences”, BIS Working Paper, No. 534, Bank for International Settlements, Basel, pp. 1-52, www.bis.org/publ/work534.pdf.

Caballero, R.J. and M.L. Hammour (1996), “On the timing and efficiency of creative destruction”, Quarterly Journal of Economics, Vol. 111/3, Oxford University Press, Oxford, pp. 805-832, https://doi.org/10.2307/2946673.

Caballero, R.J. and M.L. Hammour (1994), “The cleansing effect of recessions”, American Economic Review, Vol. 84/5, American Economic Association, Nashville, pp. 1350-1368, https://doi.org/10.1126/science.151.3712.867-a.

Caballero, R.J., T. Hoshi and A.K. Kashyap (2008), “Zombie lending and depressed restructuring in Japan”, American Economic Review, Vol. 98/5, American Economic Association, Nashville, pp. 1943-1977, https://doi.org/10.1257/aer.98.5.1943.

Davis, S.J., R.J. Faberman and J. Haltiwanger (2012), “Labor market flows in the cross section and over time”, Journal of Monetary Economics, Vol. 59/1, Elsevier, Amsterdam, pp. 1-18, https://doi.org/10.1016/j.jmoneco.2011.10.001.

Davis, S.J., R.J. Faberman and J. Haltiwanger (2006), “The flow approach to labor markets: new data sources and micro-macro links”, Journal of Economic Perspectives, Vol. 20/3, American Economic Association, Nashville, pp. 3-26, https://doi.org/10.1257/jep.20.3.3.

Davis, S.J. and J. Haltiwanger (1999), “On the driving forces behind cyclical movements in employment and job reallocation”, American Economic Review, Vol. 89/5, American Economic Association, Nashville, pp. 1234-1258, https://doi.org/10.1257/aer.89.5.1234.

Davis, S.J. and J. Haltiwanger (1992), “Gross job creation, gross job destruction, and employment reallocation”, The Quarterly Journal of Economics, Vol. 107/3, Oxford University Press, Oxford, pp. 819-863, https://doi.org/10.2307/2118365.

Davis, S.J. and J. Haltiwanger (1990), “Gross Job Creation and Destruction: Microeconomic Evidence and Macroeconomic Implications”, in O.J. Blanchard and S. Fischer (eds.), NBER Macroeconomics Annual 1990, Vol. 5, MIT Press, Cambridge, pp. 123-168, https://doi.org/10.1146/annurev-economics-061109-080430.

Foster, L., C. Grim and J. Haltiwanger (2016), “Reallocation in the Great Recession: Cleansing or not?”, in D. Card and A. Mas (eds.), Labor Markets in the Aftermath of the Great Recession, National Bureau of Economic Research, Cambridge, pp. 293-331, https://doi.org/10.1086/682397.

Fukao, K., Y.G. Kim and H.U. Kwon (2008), “Plant turnover and TFP dynamics in Japanese manufacturing”, in Chapter 3 of J.D. Lee and A. Heshmati (eds.), Micro-Evidence for the Dynamics of Industrial Evolution: The Case of the Manufacturing Industry in Japan and Korea, Nova Science Publication, New York, pp. 23-60.

Fukao, K. et al. (2011), “An international comparison of the TFP levels and the productivity convergence of Japanese, Korea, Taiwanese and Chinese listed firms”, Journal of Chinese Economic and Business Studies, Vol. 9/2, Taylor & Francis, Milton Park, pp. 127-150, https://doi.org/10.1080/14765284.2011.568683.

Fukao, K. et al. (2009), “Estimates of multifactor productivity, ICT contributions and resource reallocation effects in Japan and Korea”, RIETI Discussion Paper Series, No. 09-E-021, Research Institute of Economy, Trade and Industry, Tokyo.

Fukao, K., T. Miyagawa and M. Takizawa (2007), “Productivity growth and resource reallocation in Japan”, CEI Working Paper Series, No. 2007-9, Center for Economic Institutions, Tokyo, https://hermes-ir.lib.hit-u.ac.jp/rs/bitstream/10086/15743/1/wp2007-9a.pdf.

Hosono, K. and M. Takizawa (2015), “Misallocation and establishment dynamics”, RIETI Discussion Paper Series, No. 15-E-11, Research Institute of Economy, Trade and Industry, Tokyo.

Inui, T. et al. (2011), “Productivity dynamics and Japan’s economic growth: An empirical analysis based on the Financial Statements Statistics of Corporations by Industry”, RIETI Discussion Paper Series, No. 11-J-42, Research Institute of Economy, Trade and Industry, Tokyo.

Kwon, H.U., F. Narita and M. Narita (2015), “Resource reallocation and zombie lending in Japan in the 1990s”, Review of Economic Dynamics, Vol. 18/4, Elsevier, Amsterdam, pp. 709-732, https://doi.org/10.1016/j.red.2015.07.001.

Lucchese, M. and M. Pianta (2012), “Innovation and employment in economic cycles”, Comparative Economic Studies, Vol. 54/2, Springer, New York, pp. 341-359, https://doi.org/10.1057/ces.2012.19.

Ministry of Internal Affairs and Communications (2016), “Labour Force Survey: Historical Data”, Ministry of Internal Affairs and Communications, Japan, www.stat.go.jp/english/data/roudou/lngindex.htm (accessed 11 August 2016).

Mortensen, D.T. and C.A. Pissarides (1994), “Job Creation and Job Destruction in the Theory of Unemployment”, Review of Economic Studies, Vol. 61/3, pp. 397-415, https://doi.org/10.2307/2297896.

Osotimehin, S. and F. Pappadà, (2016), “Credit Frictions and the Cleansing Effect of Recessions”, Economic Journal, https://doi.org/10.1111/ecoj.12319.

Research Institute of Capital Formation, Development Bank of Japan (2015), Corporate Financial Data Bank (database), Development Bank of Japan, Japan, www.dbj.jp/ricf/databank/ (in Japanese).

Research Institute of Economy, Trade and Industry (2015), The Japan Industrial Productivity Database 2015 (database), Research Institute of Economy, Trade and Industry, Japan, www.rieti.go.jp/en/about/about.html (accessed 26 October 2016).

Notes

← 1. Research Institute of Capital Formation, Development Bank of Japan (2015) is used for analysis.

← 2. In non-manufacturing industries, fisheries, mining, construction and financial sectors are included. See Table 9.2 for a concrete list of non-manufacturing industries.

← 3. Because the industry year dummies are included as the control variables, the main effect of industry growth is not estimated.