Bulletin of Monetary Economics and Banking, Vol. 21, No. 2 (2018), pp. 161 - 176
DIVORCE AND HOUSING PRICE IN 31 PROVINCES
OF CHINA
Mingbo Zheng1, Yin E Chen2,
1School of Economics and Finance, Xi’an Jiaotong University, Shaanxi, China.
2School of Economics and Management, Changsha University of Science & Technology,
Changsa, China.
3Shih Chien University, Kaohsiung, Taiwan. Email: cpchang@g2.usc.edu.tw
ABSTRACT
This paper investigates the
Keyword: Divorce; Housing price; Panel cointegration test; FMOLS.
JEL Classification: D1; J12; R21.
Article history: |
|
Received |
: July 5, 2018 |
Revised |
: September 23, 2018 |
Accepted |
: September 28, 2018 |
Available online : October 31, 2018
https://doi.org/10.21098/bemp.v21i1.942
162Bulletin of Monetary Economics and Banking, Volume 21, Number 2, October 2018
I. INTRODUCTION
“Couples in one Chinese city are increasingly faking divorce to cheat property purchase restrictions, with some pretending to separate seven times to buy a new house, according to People’s Daily. … The practice of faking divorce has become widespread with many reportedly admitting the sole purpose was just to purchase a property with lower taxes.”4 China witnessed soaring housing prices, rising from 2,112 to 5,032 yuan per square meter over the period
level.
The
Many researchers hold that house demand plays an important role in determining housing price (Li and Chand, 2012; Zhang et al., 2012). One of the factors affecting housing demand is the divorce rate. A higher divorce rate may result in a higher housing price by creating more households and more housing demand (Dieleman and Schouw, 1989). Wei and Zhang (2011) propose that the imbalance of sex ratio in China forces Chinese households to save money to
4A comment published in the website of China Daily, concerning a couple’s fake divorce to cheat on housing regulations.
Divorce and Housing Price in 31 Provinces of China |
163 |
|
|
improve their children’s competitive advantages in the marriage market. Further, this gender imbalance increases housing price through higher saving rates, given that a house is recognized as a status good (Wei et al., 2017). Thus, divorce increases the demand for marriage and thus the demand for housing, implying a positive effect on housing price.
Since the pioneering work of Becker et al. (1977), the literature has documented that economic conditions, especially wealth shocks, influence marital stability (Boheim and Ermisch, 2001). Hence, Mused (2009) argues that changes in housing price significantly affect the probability of marital dissolution, which supports the proposition by Becker et al. (1977) that changes in wealth influence the decision on divorce.As Harknett and Schneider (2012) find, negative macroeconomic conditions (like recession) make couples delay the process of marital separation. Moreover, Weiss and Willis (1997), Rainer and Smith (2010) as well as Battu et al. (2013) find that housing price increases the risk of partnership dissolution. Conversely, Farnham et al. (2011) reach the reverse conclusion. Klein (2017) proposes that changes in housing prices present positive shocks to marital stability in families in the United States. For developing countries, Farzanegan and Gholipour (2016) exploit data for Iran and find that higher housing cost leads to a decline in marital stability. Fereidouni (2016) also show that there is a positive link between housing price and divorce rate in the Middle East and North Africa. Therefore, the effects of housing price on divorce remain a debatable question in the existing empirical literature. Our paper thus attempts to contribute to better understanding this issue by using the cointegration test and
As Becker et al. (1977) points out, a couple chooses to marry if the expected utility of marriage exceeds the utility of remaining single. Once the outside environment changes the gains of marriage, a couple will consider the divorce option. As Klein (2017) argues, unexpected housing price change could affect the decision of couples in the opposite direction. On the one hand, an increase in the wealth of a household (house price) means higher consumption and lower financial stress (Attanasio et al., 2011), which improves marital satisfaction and reduces the risk of divorce. On the other hand, a rapid increase in housing prices generally makes the sale of houses easier for couples on the verge of divorce (Genesove and Mayer, 1997), which increases the probability of divorce by reducing the cost of divorce and making life easier for each person (Klein, 2017).
Based on the panel provincial data, we find that there is a cointegrated relationship between housing price and divorce in the full sample. The subsample analysis also indicates that the cointegration relation exists in all subsamples. The panel VECM shows that in the
The remainder of this paper proceeds as follows. Section II describes the model and data source. Section III provides the empirical results. Finally, Section IV sets forth our conclusions.
164Bulletin of Monetary Economics and Banking, Volume 21, Number 2, October 2018
II. DATA AND MODEL
Our study uses panel data for 31 provinces in China over the period
Figure 1. Housing Price and Divorce in China
The figure shows the relationship between housing price and divorce in China
Housing Price |
|
|
|
Divorce |
7000 |
|
Housing Price |
Divorce |
3 |
6000 |
|
|
|
2.5 |
|
|
|
|
|
5000 |
|
|
|
|
|
|
|
|
2 |
4000 |
|
|
|
|
|
|
|
|
1.5 |
3000 |
|
|
|
|
2000 |
|
|
|
1 |
|
|
|
|
|
1995 |
2000 |
2005 |
2010 |
2015 |
|
|
Year |
|
|
To investigate the
(1)
Divorce and Housing Price in 31 Provinces of China |
165 |
|
|
where yit is the dependent variable, Divorce, xit represents Hprice, tr represents the time trend, αi corresponds to the
If a cointegrated relationship among the variables is determined, then one can next estimate the cointegrated vectors by using the FMOLS estimation technique (Pedroni, 2000). Furthermore, once confirming the cointegration relationship and obtaining the results, we then establish the panel VECM to investigate
(2)
(3)
The second step is to estimate the panel Granger causality model with the dynamic error correction term, as follows:
(4)
(5)
Thus, by testing the significance of the coefficients of explanatory variables in equations (4) and (5), we can identify
5Note that the cointegration test proposed by Pedroni (2004) and Westerlund (2005) has some limitations on dealing with
166Bulletin of Monetary Economics and Banking, Volume 21, Number 2, October 2018
H0:φ1k=0 for all k in equation (4) or H0:γ1k=0 for all k in equation (5). Thus, we check the significance of the speed of adjustment λ to examine
III. RESULTS
First, the panel unit root test we use are the Breitung (Breitung, 2000), the LLC (Levin et al., 2002), and the IPS (Im et al., 2003). To avoid test distortion induced by
Table 1.
Panel Unit Root Tests
The table reports the results based on the panel unit root tests. LLC, Breitung and IPS tests represent Levin et al. (2002), Breitung (2000) and Im et al. (2003) panel unit root tests, respectively, which are under the null of without a unit root. denotes first
differences. *** indicates statistical significance at the 1% level.
Variable |
LLC |
Breitung |
IPS |
|
|
|
|
Divorce |
4.007 |
6.337 |
9.852 |
Hprice |
1.689 |
0.964 |
3.778 |
ΔDivorce |
|||
ΔHprice |
|||
|
|
|
|
To correct bias induced by reverse causality and serial correlation in the ordinary least square (OLS) model, as well as to examine the
6We are grateful to the reviewer for providing this suggestion on controlling
7Eastern region: Beijing, Fujian, Guangdong, Hebei, Jiangsu, Liaoning, Shandong, Shanghai, Guangxi,
Tianjin, and Zhejiang, Hainan. Central region: Anhui, Henan, Heilongjiang, Hubei, Hunan, Jiangxi,
Inner Mongolia, Jilin, and Shanxi. Western region: Chongqing, Gansu, Guizhou, Ningxia, Shaanxi, Qinghai, Sichuan, Xinjiang, Yunnan and Tibet.
Divorce and Housing Price in 31 Provinces of China |
167 |
|
|
whether for the cointegration test of Pedroni (2004) or that of Westerlund (2005). For the eastern region, there is slightly weaker evidence based on the statistics from Pedroni (2004): only two statistics support the cointegration relationship. However, the Westerlund statistics show that the cointegration relation exists. In the central and western regions, five out of seven statistics from Pedroni (2004) reject the null of no cointegration, while the Westerlund statistics accept it. Overall, we see that housing price and divorce move together in the
Table 2.
Panel Cointegration Tests
The table reports the results of the panel cointegration tests. The test statistics are from Pedroni’s (2004) except for Westerlund that is from Westerlund (2005). Statistics are asymptotically distributed as normal. The variance ratio test is
Variable |
|
Dependent Variable: Hprice |
|
||
|
|
|
|
||
Full Sample |
Eastern |
Central |
Western |
||
|
|||||
|
|
|
|
|
|
Panel variance |
4.344*** |
1.432 |
3.485*** |
3.233*** |
|
Panel r |
0.242 |
1.289 |
|||
Panel PP |
|||||
Panel ADF |
|||||
Group r |
2.112** |
2.369** |
0.738 |
0.423 |
|
Group PP |
|||||
Group ADF |
|||||
Westerlund |
|||||
|
|
|
|
|
Table 3 gives
168Bulletin of Monetary Economics and Banking, Volume 21, Number 2, October 2018
Table 3.
FMOLS
The table reports the
Province |
Dependent Variable: |
Dependent Variable: |
|
Divorce |
Hprice |
||
|
|||
|
|
|
|
Beijing |
0.07(2.12) ** |
5.27(2.07) ** |
|
Tianjin |
0.23(5.84) *** |
3.56(5.25) *** |
|
Hebei |
0.44(7.36) *** |
2.04(8.82) *** |
|
Shanxi |
0.29(6.03) *** |
2.61(5.68) *** |
|
Inner Mongolia |
0.63(5.24) *** |
1.19(4.65) *** |
|
Liaoning |
0.50(8.97) *** |
1.61(9.43) *** |
|
Jilin |
0.92(5.59) *** |
0.80(7.18) *** |
|
Heilongjiang |
0.90(15.03) *** |
1.05(16.01) *** |
|
Shanghai |
0.06(2.96) *** |
7.64(3.03) *** |
|
Jiangsu |
0.38(10.68) *** |
2.46(9.46) *** |
|
Zhejiang |
0.21(12.40) *** |
4.49(12.40) *** |
|
Anhui |
0.49(7.12) *** |
1.79(6.18) *** |
|
Fujian |
0.25(7.59) *** |
3.36(7.10) *** |
|
Jiangxi |
0.4(8.68) *** |
2.25(9.82) *** |
|
Shandong |
0.47(26.06) *** |
2.07(26.30) *** |
|
Henan |
0.55(6.24) *** |
1.43(6.09) *** |
|
Hubei |
0.49(15.21) *** |
1.92(18.20) *** |
|
Hunan |
0.73(10.38) *** |
1.25(13.28) *** |
|
Guangdong |
0.17(10.82) *** |
5.35(10.32) *** |
|
Guangxi |
0.47(11.57) *** |
2.03(12.95) *** |
|
Hainan |
0.16(6.33) *** |
4.64(4.37) *** |
|
Chongqing |
1.00(7.45) *** |
0.79(8.68) *** |
|
Sichuan |
0.56(13.76) *** |
1.71(15.10) *** |
|
Guizhou |
0.71(8.30) *** |
1.20(7.32) *** |
|
Yunnan |
0.4(25.37) *** |
2.44(26.42) *** |
|
Tibet |
0.04(1.43) |
5.40(2.78) *** |
|
Shaanxi |
0.33(6.17) *** |
2.22(5.19) *** |
|
Gansu |
0.28(8.54) *** |
3.08(8.76) *** |
|
Qinghai |
0.21(3.63) *** |
2.82(4.00) *** |
|
Ningxia |
0.63(7.45) *** |
1.31(7.59) *** |
|
Xinjiang |
0.45(1.53) |
0.51(2.05) ** |
|
Panel |
0.43(49.55)*** |
2.59( 51.45)*** |
|
Eastern |
0.28(32.54) *** |
3.71(32.18) *** |
|
Central |
0.60(26.51) *** |
1.59(26.02) *** |
|
Western |
0.46(26.44) *** |
2.15(27.79) *** |
|
|
|
|
Divorce and Housing Price in 31 Provinces of China |
169 |
|
|
Table 4 provides results of the panel causality test. In the
Table 4.
Panel Causality Tests
The table reports the results from the panel causality tests. ⋅ denotes there is no causal relationship, and + denotes there exists a positive causal relationship. Hprice denotes housing price. ***, **, and * denote statistical significance at the 1%, 5%, and 10% levels, respectively. The parameters λ are the
|
Dependent |
Source of Causation (Divorce or Hprice) |
|||
|
Variable |
Full |
Eastern |
Central |
Western |
∆Divorce |
+ |
+ |
+ |
x |
|
|
∆Hprice |
x |
x |
x |
x |
λ |
∆Divorce |
0.863 |
2.715*** |
0.113 |
|
|
∆Hprice |
8.619*** |
4.875*** |
0.686 |
|
∆Divorce |
+ |
+ |
+ |
x |
|
|
∆Hprice |
+ |
+ |
x |
+ |
The above results confirm the cointegration relationship between housing price and divorce in 31 provinces of China. To enhance the reliability of our results, we further use residential housing price per square meter (RHprice) as the proxy variable to
Table 5 presents the results of a panel unit root test for RHprice. It is clear that RHprice exhibits a significant I(1) process, showing that RHprice and Divorce follow similar paths. Using these results, we thus test for RHprice and Divorce to determine whether there is a
170Bulletin of Monetary Economics and Banking, Volume 21, Number 2, October 2018
is supported, we use FMOLS to determine the specific influence. Table 7 gives the robust
Table 5.
Panel Unit Root Tests for RHprice
The table reports the panel unit root tests of RHprice. LLC, Breitung and IPS tests represent Levin et al. (2002), Breitung (2000) and Im et al. (2003) panel unit root tests, respectively, which are under the null of without a unit root. denotes first differences. ***
indicates statistical significance at the 1% level.
Variable |
LLC |
Breitung |
IPS |
|
|
|
|
RHprice |
4.78 |
7.412 |
11.442 |
ΔRHprice |
|||
|
|
|
|
Table 6.
Additional Panel Cointegration Tests
The table shows results of the additional panel cointegration tests for robustness. These statistics are from Pedroni’s (2004) except for Westerlund that is from Westerlund (2005). Statistics are asymptotically distributed as normal. The variance ratio test is
Variable |
|
Dependent Variable: RHprice |
|
||
Full Sample |
Eastern |
Central |
Western |
||
|
|||||
Panel variance |
1.232 |
1.352 |
3.080*** |
2.909*** |
|
Panel r |
1.308 |
1.367 |
|||
Panel PP |
|||||
Panel ADF |
|||||
Group r |
3.170*** |
2.433** |
0.819 |
0.995 |
|
Group PP |
|||||
Group ADF |
|||||
Westerlund |
Divorce and Housing Price in 31 Provinces of China |
171 |
|
|
Table 7.
Additional FMOLS
The table shows the additional FMOLS
Province |
Dependent Variable: |
Dependent Variable: |
|
Divorce |
RHprice |
||
|
|||
|
|
|
|
Beijing |
0.08(2.33) ** |
5.45(2.30) ** |
|
Tianjin |
0.24(5.94) *** |
3.40(5.38) *** |
|
Hebei |
0.44(7.87) *** |
2.07(9.25) *** |
|
Shanxi |
0.32(7.26) *** |
2.58(6.84) *** |
|
Inner Mogolia |
0.69(5.64) *** |
1.13(4.94) *** |
|
Liaoning |
0.51(9.43) *** |
1.59(9.66) *** |
|
Jilin |
0.93(5.74) *** |
0.80(7.21) *** |
|
Heilongjiang |
0.91(13.57) *** |
1.02(14.13) *** |
|
Shanghai |
0.06(3.07) *** |
8.26(3.26) *** |
|
Jiangsu |
0.39(12.03) *** |
2.44(10.92) *** |
|
Zhejiang |
0.21(11.35) *** |
4.64(11.48) *** |
|
Anhui |
0.50(7.00) *** |
1.76(6.03) *** |
|
Fujian |
0.23(6.82) *** |
3.46(6.25) *** |
|
Jiangxi |
0.40(9.53) *** |
2.27(10.68) *** |
|
Shandong |
0.49(24.39) *** |
1.98(23.99) *** |
|
Henan |
0.60(6.27) *** |
1.33(6.31) *** |
|
Hubei |
0.49(17.96) *** |
1.93(22.38) *** |
|
Hunan |
0.78(11.87) *** |
1.18(15.62) *** |
|
Guangdong |
0.17(10.45) *** |
5.42(10.07) *** |
|
Guangxi |
0.48(12.7) *** |
1.97(13.93) *** |
|
Hainan |
0.15(5.97) *** |
4.57(4.07) *** |
|
Chongqing |
1.01(10.04) *** |
0.83(11.27) *** |
|
Sichuan |
0.56(11.82) *** |
1.72(11.81) *** |
|
Guizhou |
0.73(5.18) *** |
0.96(4.17) *** |
|
Yunnan |
0.45(22.39) *** |
2.14(24.03) *** |
|
Tibet |
0.03(1.19) *** |
5.14(2.55) *** |
|
Shaanxi |
0.35(5.98) *** |
2.12(4.96) *** |
|
Gansu |
0.29(6.70) *** |
2.91(6.87) *** |
|
Qinghai |
0.22(2.54) ** |
1.87(2.49) ** |
|
Ningxia |
0.67(9.07) *** |
1.27(8.93) *** |
|
Xinjiang |
0.50(1.68) * |
0.56(2.77) *** |
|
Panel |
0.45(49.18)*** |
2.54( 51.11)*** |
|
Eastern |
0.29(32.43) *** |
3.77(31.92) *** |
|
Central |
0.63(28.29) *** |
1.56(31.38) *** |
|
Western |
0.48(24.22) *** |
1.95(25.25) *** |
|
|
|
|
172Bulletin of Monetary Economics and Banking, Volume 21, Number 2, October 2018
Table 8 provides results from the panel causality test. For the
Table 8.
Additional Panel Causality Tests
The table shows the additional panel causality tests. ⋅ denotes there is no causal relationship, and + denotes there exists a positive causal relationship. RHprice denotes residential housing price. ***, **, and * denote statistical significance at the 1%, 5%, and 10% levels, respectively. The parameters λ are the
|
Dependent |
Source of Causation (Divorce or RHprice) |
|||
|
Variable |
Full |
Eastern |
Central |
Western |
∆Divorce |
+ |
+ |
x |
x |
|
|
∆RHprice |
x |
x |
x |
x |
λ |
∆Divorce |
0.834 |
2.632*** |
0.213 |
|
|
∆RHprice |
10.084*** |
5.672*** |
1.124 |
|
∆Divorce |
+ |
+ |
x |
x |
|
|
∆RHprice |
+ |
+ |
x |
x |
IV. CONCLUSION
Employing data for 31 Chinese provinces over the period
Overall, our results confirm the cointegration relationship between housing price and divorce, specifically
Divorce and Housing Price in 31 Provinces of China |
173 |
|
|
the
REFERENCES
Attanasio, O., Leicester, A., and Wakefield, M. (2011). Do House Prices Drive Consumption Growth? The Coincident Cycles of House Prices and Consumption in the UK. Journal of the European Economic Association, 9,
Battu, H., Brown, H., and
Becker, G. S., Landes, E. M., and Michael, R. T. (1977). An Economic Analysis of Marital Instability. Journal of Political Economy, 85,
Boheim, R., and Ermisch, J. (2001). Partnership Dissolution in the
Breitung, J. (2000). The Local Power of Some Unit Root Tests for Panel Data. Advances in Econometrics, 15,
Chang, C., and Lee, C. (2010). A
Chang, C., and Lee, C. (2015). Do Oil Spot and Futures Prices Move Together?. Energy Economics, 50,
Chang, C., Lee, C., and Berdiev, A. (2015). The Impact of Government Ideology on Energy Efficiency: Evidence from Panel Data. Energy Efficiency, 8,
Chang, C., Lee, C., and Hsieh, M. (2011). Globalization, Real Output, and Multiple Structural Breaks. Global Economic Review, 40,
Dieleman, F. M., and Schouw, R. J. (1989). Divorce, Mobility and Housing Demand. European Journal of Population, 5,
Du, Z., and Zhang, L. (2015).
Engle, R. F., and Granger, C. W. J. (1987).
Fan, Z., and Hu, X. (2015). Unexpected
Farnham, M., Schmidt, L., and Sevak, P. (2011). House Prices and Marital Stability. The American Economic Review, 101,
Farzanegan, M. R., and Gholipour, H. F. (2016). Divorce and the Cost of Housing: Evidence from Iran. Review of Economics of the Household, 14,
174Bulletin of Monetary Economics and Banking, Volume 21, Number 2, October 2018
Fereidouni, H. G. (2016). Housing Costs and Divorce Rate in the MENA Countries. Topics in Middle Eastern and North African Economies, 18.
Genesove, D., and Mayer, C. J. (1997). Equity and Time to Sale in the Real Estate Market. The American Economic Review, 87,
Harknett, K., and Schneider, D. (2012). Is a Bad Economy Good for Marriage? The Relationship between Macroeconomic Conditions and Marital Stability from
Im, K. S., Pesaran, M. H., and Shin, Y. (2003). Testing for Unit Roots in Heterogeneous Panels. Journal of Econometrics, 115,
Jang, C., and Chang, C. (2014). National Income and Fishery Consumption: A Global Investigation. Economic Research - Ekonomska Istraživanja, 27,
Klein, J. (2017). House Price Shocks and Individual Divorce Risk in the United States. Journal of Family & Economic Issues, 38,
Levin, A. T., Lin, C. and Chu, C. J. (2002). Unit Root Tests in Panel Data: Asymptotic and
Li, L., and Wu, X. (2014). Housing Price and Entrepreneurship in China. Journal of Comparative Economics, 42,
Li, P., and Xu, J. (2012). Housing Price and Fertility Rate. China Economic Journal, 5,
Li, Q., and Chand, S. (2013). House Prices and Market Fundamentals in Urban China. Habitat International, 40,
Li, S., Whalley, J. and Zhao, X. (2013). Housing Price and Household Savings Rates: Evidence from China. Journal of Chinese Economic and Business Studies, 11,
Liang, W., Lu, M., and Zhang, H. (2016). Housing Prices Raise Wages: Estimating the Unexpected Effects of Land Supply Regulation in China. Journal of Housing Economics, 33,
Mused, A. (2009). Another Look at Wealth and Marital Relationships:The Effects of House Prices on Divorce Rates. Working Paper,
Pedroni, P. (2000). Fully Modified OLS for Heterogeneous Cointegrated Panels. Advances in Econometrics, 15,
Pedroni, P. (2004). Panel Cointegration: Asymptotic and Finite Sample Properties of Pooled Time Series Tests with an Application to the PPP Hypothesis. Econometric Theory, 20,
Peteke, F., and Maarten, V. H. (2010). The Impact of Splitting up and Divorce on Housing Careers in the UK. Housing Studies, 25,
Rainer, H., and Smith, I. (2010). Staying Together for the Sake of the Home? House Price Shocks and Partnership Dissolution in the UK. Journal of The Royal Statistical Society Series
Wang, Q., and Zhou, Q. (2010). China’s Divorce and Remarriage Rates: Trends and Regional Disparities. Journal of Divorce & Remarriage, 51,
Wang, X. (2018). Divorce and Housing Price
Wen, J., Yang, X., Feng, G., Sui, B., and Chang, C.P. (2017). The Comovement Between Venture Capital and Innovation in China: What Are The Implications?. Quality & Quantity, 51,
Divorce and Housing Price in 31 Provinces of China |
175 |
|
|
Wei, S. and Zhang, X. (2011). The Competitive Saving Motive: Evidence from Rising Sex Ratios and Savings Rates in China. Journal of Political Economy, 119,
Wei, S., Zhang, X., and Liu, Y. (2017). Home Ownership as Status Competition: Some Theory and Evidence. Journal of Development Economics, 127,
Weiss, Y., and Willis, R. J. (1997). Match Quality, New Information, and Marital Dissolution. Journal of Labor Economics, 15,
Westerlund, J. (2005). New Simple Tests for Panel Cointegration. Econometric Reviews, 24,
Yu, W., and Zhou, W. (2015). The Research on Housing Price and Marital Stability in China. Statistics & Decision, 24,
Zhang, C., An, G., and Yu, X. (2012). What Drives China’s House Prices: Marriage or Money?. China & World Economy, 20,
176Bulletin of Monetary Economics and Banking, Volume 21, Number 2, October 2018
Appendix
Table A1.
Descriptive Statistics
The table shows the descriptive statistics of the variables, Divorce and Hprice, for the period of 1997 to 2015. Divorce and Hprice denote divorce and housing price, respectively. Min and Max denote minimum and maximum, respectively.
Variables |
Observation |
Mean |
Standard |
Min |
Max |
|
Deviation |
||||||
|
|
|
|
|
||
|
|
|
|
|
|
|
Divorce |
589 |
6.855 |
5.335 |
0.050 |
28.620 |
|
Hprice |
589 |
3.581 |
2.972 |
0.138 |
22.633 |
|
|
|
|
|
|
|