|Interviewee||ITO Arata (Fellow, RIETI)|
|Issue date / NO.||Research Digest No.15|
|Download / Links|
There is "a lack of clarity" and "uncertainty" over various government policies, e.g., re-postponement of the consumption tax rate increase, negotiations for the Trans Pacific Partnership (TPP), and monetary policy. How would this affect consumption activities and investment activities? RIETI Fellow Arata Ito prepared the News-based Economic Policy Uncertainty Index for Japan based on newspaper articles to quantify the uncertainty of policy steering, so as to analyze the relationship between policy uncertainty and the economy. The study has found that the increase in policy uncertainty is the pre-stage of decline in economic activities. The finding suggests that clarifying future policy steering to reduce policy uncertainty could prevent deterioration of economic performance. Ito hopes that the index will prove to be useful for conducting economic research at both the government and private levels, and implemented in academic studies for accumulating insight into policy uncertainty.
Research background and methodology
—Tell us about your motivation behind conducting this study.
The world financial crisis from 2007 to 2008 caused the economy to recede in many countries. The economy has since recovered, but the pace of recovery has been quite slow. Why is it that the economic recovery has not gained momentum? Overseas academics and media have presented several theories. One of them is with regard to policy uncertainty. For example, Stanford University Professor John Taylor cited policy uncertainty as one of the main inhibitors of economic recovery. He notes that—amidst the poor outlook of the medical system reform, financial system reform, as well as the monetary policy of the Federal Reserve Board (FRB)—reduced policy uncertainty would encourage companies to utilize their surplus fund, thereby stimulating the economy.
Researchers keenly observed whether actual data would support this theory. A team of researchers including University of Chicago Professor Steven Davis and Stanford University Professor Nick Bloom developed a new indicator called the Economic Policy Uncertainty Index, focusing on newspaper coverage, as I will explain later. Today, substantial knowledge has been accumulated based on this index. Yet, Japan's level of knowledge on policy uncertainty is quite poor compared to that of European and North American countries. As the first step for narrowing the gap, I thought it was necessary to ask three basic questions, i.e., how Japan's policy uncertainty has changed with time, which policy areas are the source of increasing uncertainty, and how does policy uncertainty relate to economic activities.
—Tell us about actual results provided in your paper, particularly about policy implications.
This policy discussion paper provides policy practitioners with the results of joint research by Professor Davis, economists at the International Monetary Fund's (IMF) Regional Office for Asia and the Pacific, and myself. It gives three main findings, corresponding to the three questions mentioned earlier. First of all, the policy uncertainty index rose when a new prime minister was inaugurated or at the time of fiercely-contested national elections. The index also rose at the Asian currency crisis from 1997 to 1998, collapse of Lehman Brothers in 2008, U.S. debt ceiling crisis and U.S. credit rating downgrade in 2011, and, more recently, the re-postponement of the consumption tax rate increase in 2016. As you can see, Japan's policy uncertainty rises in response to not only domestic factors but also overseas factors.
Next, the study also found that the rise in policy uncertainty is attributable primarily to factors associated with fiscal policies. More specifically, about 60% of all factors are related to fiscal policies, while about 30% are associated with monetary policies. Finally, simple quantitative analysis shows that the increase of policy uncertainty is a precursor of decline in economic performance.
Third, the study has indicated that the economy benefits from having policy uncertainty kept low by way of stabilizing the government's policy steering and clarifying policy outlook.
—Can you explain the method for measuring policy uncertainty?
There are several approaches to measuring policy uncertainty. The first approach is to conduct a survey. The advantage of this approach is that it directly captures uncertainty faced by households and companies of varying characteristics. Its disadvantage is the amount of work involved in conducting a large-scale survey at high frequency, e.g., weekly or monthly. It is also difficult to obtain retrospective data, making this approach unsuitable for time series analysis.
There is also an approach of using newspaper coverage. The research group led by Professor Bloom and Professor Davis developed an indicator of policy uncertainty by focusing on the frequency of newspaper coverage related to uncertainty. Against the backdrop of this approach lies an assumption that households and companies must be facing a state of high uncertainty when newspapers carry a large number of such articles. The advantage of this approach is the retrospective availability of high coverage data. Databases of articles, held by major newspapers, can be used to search relevant articles published over a long period of time. Its disadvantage is the difficulty in capturing uncertainty faced by households and companies of varying characteristics. This study adopted the latter approach as it caters to the purpose of the study.
Methodology of the News-based Economic Policy Uncertainty Index for Japan
—Can you provide details of the methodology for compiling the News-based Economic Policy Uncertainty Index for Japan?
Of all of the articles published in the Nikkei, Yomiuri, Asahi, and Mainichi newspapers, I searched for those that contain at least one term relating to the three categories listed below for each month. The article search covered the period from January 1987 to today.
Economy: "keizai (economy)," "keiki (economic condition)"
Uncertainty: "futomei (unclear)," "fukakujitsu (uncertain)," "fukakutei (unconfirmed)," "fuan (concern)"
Policy: "zeisei (taxation system)," "saishutsu (expenditures)" and other policy-related terms
Japanese terms corresponding to "economy" and "uncertainty" have been chosen through careful comparison of Japanese articles and their English translations provided by the respective newspaper publishers. Let's take the case of "uncertainty" for example. First, I randomly selected a specific number of articles from the searched articles containing the words "uncertainty" or "uncertain." Then, the selected articles were compared against their Japanese equivalent articles to identify Japanese terms corresponding to "uncertainty" or "uncertain." This found that "futomei," "fuan," "bimyo," "fukakujitsu," "fuantei," and "fukakutei" were used as corresponding terms at a relatively high frequency.
Then, I randomly selected a specific number of articles from articles containing both "keizai" or "futomei." The selected articles were compared against their English equivalent articles to identify English terms corresponding to "futomei." The same comparison was made for the other five Japanese terms mentioned above as well. Of the six terms, those that corresponded to "uncertainty" or "uncertain" at a high frequency were "futomei," "fuan," "fukakujitsu" and "fukakutei." This study therefore adopted these four terms as Japanese equivalents of "uncertainty." The same process was applied to the term "economy," and the study consequently adopted "keizai" and "keiki" as its Japanese equivalents.
As for the "Policy" category, I used many of the terms that Professor Davis et al. listed as candidate terms when they compiled the U.S. version of the index. The aforementioned method was used to determine corresponding Japanese terms. In the end, I adopted a total of 32 terms including "zeisei (taxation system)," "seifu saimu (government debt)," "kisei (regulations)," "hoan (legislative bill)," and "Nihon Ginko (Bank of Japan)."
Data of newspaper coverage, collected this way, were applied with seasonal adjustment and other treatments, before calculating an index with the average for January 1987–December 2015 set at 100. Indices were also prepared for individual policies, in addition to the index for all policies. More specifically, indices were compiled for fiscal policy, monetary policy, foreign exchange policy, and trade policy. Of all of the articles collected in compiling the index for all policies, those containing each of the individual policies were selected to calculate the respective indices. Index data are compiled monthly, starting from January 1987.
—Tell us more about the characteristics of the pre-2000 index.
This figure shows the transition of the News-based Economic Policy Uncertainty Index for Japan. The higher the index is, the greater is the level of uncertainty. Until the mid 1990s, the index hovered around just below 100 in general. However, there were times when the index showed a sharp spike, e.g., in October 1987. This was when Japan's political landscape became destabilized over the choice of successor to Prime Minister Yasuhiro Nakasone. Meanwhile, overseas, U.S. President Ronald Reagan was in fierce confrontation with the U.S. Congress over the reduction of fiscal deficits.
One of the periods when the index reached a very high level over the last 30 years was the late 1990s. At the time, the Cabinet led by Prime Minister Ryutaro Hashimoto adopted austere fiscal management to rebuild fiscal health. However, the 1997 Asian currency crisis caused the economy to rapidly deteriorate. In response, Japan's political opposition and even some forces within the ruling party began calling for shelving the fiscal reconstruction initiative and shifting into a more aggressive fiscal policy. The standoff between those promoting fiscal reconstruction and those seeking an aggressive fiscal approach heightened uncertainty over fiscal policy.
The index rose again in 1998. In July, the Liberal Democratic Party (LDP) lost its majority in the Upper House election, allowing the opposition camp to claim majority in the chamber. The sense of concerns rose over the possible difficulty in policy implementation under this so-called "Twisted Diet." The LDP then forged a coalition with the Liberal Party, etc. to regain Upper House majority, resolving the "twisted" state of the Diet. This caused the index to drop sharply.
—Can you explain details about post-2000 periods when the index rose?
As you can see in the figure, the index has spiked frequently since 2000, e.g., in 2001. Then, the resignation of Prime Minister Yoshiro Mori triggered infighting within the ruling party, driving up political uncertainties. With no further room to lower the policy rate, there were discussions about the Bank of Japan (BoJ) introducing new monetary easing measures. Fast forward to 2008, the index reached the highest level over the last 30 years. Discussions were made on what policy approach the government should take, especially in terms of fiscal policy, to address economic recession triggered by the world financial crisis. In the current decade, the index spiked at major events such as the European sovereign crisis, the U.S. increase of federal debt ceiling, the BoJ's decision to adopt a negative interest rate policy, and the re-postponement of consumption tax increase. Most recently, the index rose sharply at the time of the U.S. presidential election in November 2016.
—The Shinzo Abe administration's decision to re-postpone the consumption tax rate increase has significantly heightened uncertainty. How does the consumption tax rate increase relate to uncertainty?
First, it is a good idea to examine how the fiscal policy's uncertainty index reacted when it was decided to postpone the increase in consumption tax rate. At the time of first postponement, i.e., in November 2014, the index rose by about 20 points. At the time of the second postponement, i.e., in May 2016, the index rose by about 80 points. This indicates that the postponement of consumption tax rate increase does not necessarily heighten uncertainty. Why was it that the second postponement raised uncertainty to that extent? I believe it was attributable to the prime minister's failure to present convincing and specific preparation for clearing the fiscal consolidation goals to which he had committed. In a news conference announcing the re-postponement, the prime minister indicated that the increase in tax revenues, resulting from economic growth, would be partially used to meet the rising fiscal demand. However, he fell short of clearly stating how he planned to secure the tax rate increase in 2019, and generate fiscal surplus until then.
Uncertainty of monetary and trade policies
—Tell us about the uncertainty of monetary and trade policies, and how the recent TPP development and U.S. presidential election relates to uncertainty.
Let me talk about the monetary policy's uncertainty index first, and then about trade policy's uncertainty index. The monetary policy uncertainty index has three characteristics. First, the index rises just before and after any changes to the framework of monetary policy, e.g., in early 2001. At the time, with little room left for lowering the policy rate, there were debates on the introduction of new monetary easing measures. In March, the BoJ adopted the quantitative easing policy, in which the outstanding balance of current accounts held by Japanese banks at the BoJ was used as the main target for monetary operations in the financial market. The index also spiked at the start of 2016, when the BoJ adopted the negative interest rate policy. The index's second characteristic is its sharp increase when the Japanese yen's exchange rate reaches around 80 to the U.S. dollar. It rose in 1995 amidst debate on the official discount rate cut, and in 2010-2011 during the debate on expanding the qualitative easing policy. The third characteristic of the index was seen in early 2008, when it rose due to fear that the position of BoJ governor could become vacant. This was because the Twisted Diet made it difficult for the ruling party to pass BoJ governor nominations the government had submitted.
As for the trade policy's uncertainty index, it showed significant increases twice in the past. The first surge came in the period from the late 1980s to the early 1990s. This includes 1988, when the U.S. Congress deliberated on the bill for the Omnibus Foreign Trade and Competitiveness Act, and 1993, when final trade negotiations for an agreement were held in the General Agreement on Tariffs and Trade (GATT)'s Uruguay Round. The other increase came in the period from 2011 to present. The index's fluctuations are closely associated with developments surrounding the TPP Agreement. For example, the index spiked in November 2011, when the ruling Democratic Party of Japan was suffering from severe infighting over whether Japan should participate in TPP negotiations. The index also rose from 2014 into 2015, when the level of concerns built up over the final agreement on TPP negotiations and its parliamentary ratification procedure. The index reached its highest level over the last 30 years in November 2016 at the time of the U.S. presidential election, and in January 2017 at the inauguration of the new U.S. President Donald Trump. This was because of the significant uncertainty created by the U.S. decision to leave the TPP Agreement
Further development and future tasks
—How do you plan to further expand this study, which includes this paper as well as your previous papers?
The index I prepared this time has several issues, including the selection of terms in the "Policy" category. It is necessary to more carefully examine terms that appear frequently in articles that refer to policy uncertainty, as the U.S. index does. This paper could not accomplish it.
This study compiled a new index based on newspaper coverage in order to answer three basic questions about Japan's policy uncertainty. I hope this index will have proved to be useful for conducting economic research at both the government and private levels, and be utilized in academic studies for accumulating knowledge about policy uncertainty.
Arata Ito has been a fellow at RIETI since 2013. After receiving his M.A. from Hitotsubashi University, he became a research associate at the Institute of Economic Research, Hitotsubashi University in 2009. In 2012 he joined the Graduate School of Economics, University of Tokyo as a project academic support specialist. His recent works include: "Fiscal policy switching in Japan, the U.S., and the U.K.," Journal of The Japanese and International Economics, 25 (4), 2011 (with Tsutomu Watanabe and Tomoyoshi Yabu) and "Policy Uncertainty in Japan," NBER Working Paper No. 23411, 2017 (with Elif C. Arbatli, Steven J. Davis, Naoko Miake, Ikuo Saito)