BEIJING — With economic growth slowing and China’s stock markets in turmoil, the Communist Party and central government are starting to restrict the flow of economic information and data to control the narrative. That runs counter to the needs of global investors and economists, who had been hoping that China would continue becoming more transparent with such information.

北京——随着经济增长放缓、中国股市处于动荡之中，中国共产党和中央政府正在开始用限制经济信息和数据的流动来控制话语权。这种做法与全球投资者和经济学家所需要的背道而驰，他们一直在希望，中国在这些信息上会继续变得更加透明。

In 2007, Li Keqiang, who has since become China’s premier, told the United States ambassador in Beijing about the unreliability of official economic growth data. Among people tracking the Chinese economy, that kind of doubt is likely to increase. The party is becoming more sensitive as to how data shapes global perceptions of the economy.

2007年，后来出任中国总理的李克强曾在北京对美国驻华大使说官方经济增长数据不可靠。在跟踪中国经济的人们中，对经济数据的怀疑很可能会增加。共产党正在对数据如何塑造全球对中国经济的认识变得更敏感。

The New York Times has found prominent recent cases of data and information censorship. Last year, officials at China’s National Bureau of Statistics ordered Markit Economics, a British financial information and services company, and Caixin Media Group, based in Beijing, to end a monthly release of preliminary survey results of Chinese factories. Because traders used the number, called the flash P.M.I., the data affected markets.

《纽约时报》已发现了几起最近的数据和信息审查的突出案例。去年，中国国家统计局官员曾命令英国金融信息和服务公司Markit Economics和总部设在北京的财新传媒集团，停止每月发布中国工厂调查的初步结果。因为交易者使用这个被称作PMI预览指数的数据，而这个数据对股市有影响。

Here are comments from six experts on the quality and official control of Chinese economic data:

以下是六位专家对中国经济数据质量及官方控制问题的意见：

Ye Tan, independent economist based in Shanghai

叶檀，驻上海独立经济学家

We know that every institution or agency has its own approach. So we have a lot of data, but it’s kind of messy.

我们知道，每个机构或部门都有自己的方法。所以，我们有大量的数据，但这些数据很乱。

Some economic data like income and unemployment is indeed sensitive in China. But the authorities will have no choice but to release them, or else they will be creating a situation where a blind man is trying to feel an elephant.

像收入和失业率这些经济数据在中国确实很敏感。不过，当局需要发布这些数据，别无选择，否则他们会制造一种盲人摸象的局面。

Christopher Balding, associate professor at Peking University HSBC Business School

克里斯托弗·鲍尔丁(Christopher Balding)，北京大学汇丰商学院副教授

Many parts of Chinese data are riddles inside enigmas, so I’ve learned to never take anything as final.

中国数据的许多部分是谜中之谜，所以我学会了从不把任何东西当作最终结果。

Chinese data even among specialists are hazy and require time just sorting through the definitions and changes.

即使对专家们来说，中国的数据也是模糊的，仅仅把定义和变化搞清楚就需要大量的时间。

I don’t think anyone has any real illusions that the economy is growing at 7 percent. Only a couple of people would write that with a straight face anymore. I think the party is becoming a lot more sensitive to that.

我认为没有任何人对经济以7％的速度增长抱有任何真正的幻想。只有少数人会绷着脸继续这样写。我认为共产党对这一点变得更加敏感了。

The People’s Bank of China has to learn how to interact with the markets rather than just repeating “everything is fine,” because people are watching a lot closer than they used to.

中国人民银行必须学会如何与市场互动，而不是不断地重复“一切都好”，因为人们正在更近的距离上观察，比以前的近多了。

China used to publish widespread aggregate employment levels throughout secondary industry [the manufacture of finished products]. In 2013, they stopped publishing the aggregate number of employed persons in these industries. They just started releasing a couple of aggregate employment levels for small numbers of subsectors. It’s very limited, but it’s a start.

中国曾发布整个第二产业（商品制造业）范围广泛的总体就业水平数据。2013年，政府停止了发布在这些行业中就业人员的总数。最近才刚开始发布少量细分行业的总体就业水平。尽管十分有限，但是一个开端。

They haven’t published wage data since the end of 2014.

政府自2014年底以来一直都没有公布过工资数据。

Yuan Gangming, researcher at the Center for China in the World Economy, Tsinghua University

袁钢明，清华大学中国与世界经济研究中心研究员

The National Bureau of Statistics is not as unreliable as it used to be. I’ve been following Chinese economic data for quite some time. In the past, the forgery of data was common. In 1998, during the Asia financial crisis, some data like industrial output and electricity were not made available to the public.

国家统计局并不像以前那样不可靠。我关注中国经济数据有一段时间了。在过去，伪造数据的情况很常见。1998年亚洲金融危机期间，没有提供给公众像工业产出和用电量这样的数据。

Over all, the National Bureau of Statistics is a lot more reliable than some local statistics bureaus.

总的来说，国家统计局比一些地方上的统计局要可靠得多。

Anne Stevenson-Yang, co-founder of J Capital Research

杨思安(Anne Stevenson-Yang)，美奇金(J Capital Research)共同创始人

Senior officials require that the various institutions that collect the data be in on the original sources and ask those sources to change the data when needed.

高级官员要求收集数据的不同机构知道原始数据的来源，并在有必要时，要求这些来源更改数据。

There’s a small group at the National Bureau of Statistics that takes in the provincial numbers for gross domestic product, and when these conflict with national numbers, they make a judgment call.

国家统计局有一个从各省收集国内生产总值数据的小组，当各省的数字与国家的数字有冲突时，他们要做出自己的判断。

Data disappears when it becomes negative. That’s happened in a lot of series. The Ministry of Commerce has a series on retail enterprises and retail sales of various products. Those series stop when the numbers get bad.

如果数据变成负值的话，就会消失。这在很多数据序列都发生过。商务部有零售企业及各种商品的零售销售额的序列。如果一些数字不好的话，这些序列就会停下来。

I’ve been trying to deal with household income data. That’s frustrating. They report that in different ways all the time, so the numbers aren’t comparable.

我一直在试图处理家庭收入数据。但很令人沮丧。他们总是以不同的方式发布数字，所以这些数字没有可比性。

The most egregious faking is in services and job creation.

最令人震惊的作假是在服务业和创造就业机会的数字上。

For steel numbers, they report the steel output of a select number of mills. They randomly add different companies to the sample that had never been there before. We found that happening in 2014. They added all these phantom companies that nobody had heard of before. The companies report a raw number of tons of steel produced. There were 85 companies. Then suddenly there were 95 companies. They don’t tell you about this change upfront. They’ll only tell you if you ask them.

至于钢产量的数据，他们报告的是有选择的几家钢厂的钢产量。他们在样本中随便加入从来都不存在的其他公司。我们在2014年发现了这个问题。他们加进去了所有这些以前没有人听说过的幽灵公司。每家公司报上来的是钢产量的原始吨数。有85家公司报了数据。然后突然变成了95家公司。他们不把这个变化坦率地告诉你。只是在你问的时候，他们才告诉你。

On retail data, they change the companies that are the key enterprises. They don’t consistently define key enterprises. Usually you can find the definitions on the National Bureau of Statistics website. But those change. This year, a key enterprise might be defined as over 25 million renminbi per year in revenue.

在零售数据方面，他们会对哪些公司属于重点企业进行修改。他们对重点企业的定义缺乏连贯性。通常你可以在国家统计局的网站上找到定义，但这个定义是会变化的。今年，重点企业可能会被定义为每年营收超过2500万元人民币的企业。

On housing price data, the localities change different districts of their cities for the basis of comparison. If they think their prices are too high, they’ll choose a district of the city that has a high average price. Then they’ll use that as the basis of comparison. If you want low year-on-year growth, you need a high basis of comparison. If you want growth to look high, take down the basis of comparison.

在房价数据上，各地会用市内不同的区域作为比较基础。如果他们认为本市房价太高，就会选择均价较高的一个区，把它作为比较的基础。如果你想让同比增长率看起来比较低，就需要用高一些的比较基础。如果你想让增长率看起来高一些，就要把比较的基础放低一些。

Trucking is a very fragmented world in China. There are a zillion tiny companies that only weigh one truck or two trucks. There are some stations that weigh trucks. But it’s hard for me to believe that anyone reports accurate numbers on these trucks.

在中国，卡车运载量的数据非常零碎。无数个微型公司只对一、两辆卡车进行称重。有一些站点会对卡车称重。但是在我看来，任何人统计的卡车载重数字，其准确度都很难相信。

Mark Williams, chief Asia economist for Capital Economics

马克·威廉姆斯(Mark Williams)，凯投宏观(Capital Economics)首席亚洲经济学家

Unfortunately, it looks increasingly likely that the government will continue to announce G.D.P. growth targets. That will include a target of average growth over the five-year period. The government appears to view targets as essential for policy making. It is unlikely that the National Bureau of Statistics would publish numbers lower than the target number.

糟糕的是，中国政府继续发布GDP增长目标的可能性似乎提高了。这包括五年内的平均增长目标。政府似乎把制定目标看成是政策制定中至关重要的一环。国家统计局不太可能公布低于目标的数字。

China has set G.D.P. growth targets for a long time. For most of the past 20 years, it really didn’t matter that much because the economy was growing much faster than the targets. It has only really become a problem as growth has slowed.

中国设定GDP增长目标已经有很长一段时间了。在过去20年里的大多数时候，这都不是太碍事，因为经济的增长速度比目标高得多。只有在增长有所放缓的时候，它才真正成为了问题。

There’s also much more scrutiny on policy makers for their economic policy making. It wasn’t that long ago that the world felt the Chinese government could do no wrong when it came to managing their economy. But there’s a lot more skepticism now both in China and abroad.

决策者对经济政策的制定也受到了远更严格的审视。不久之前，世界还觉得中国政府在管理本国经济方面不会出错。但现在，无论国内还是国外，人们的怀疑都大大加剧了。

We used to have debates with clients on whether growth was as slow as what our proxy measures suggested. In the past year or so, I haven’t had anyone dispute that growth is significantly lower than official figures. Most investors take it as a given that the official figures are overstating China’s growth.

之前，对于中国经济增长是否像我们的代理数据显示的那样缓慢，我们常常与客户发生争论。但在过去一年左右的时间里，还没有任何人和我争论，中国经济增长是否比官方数字低很多。多数投资者都把中国官方夸大经济增长看成了一种常态。

The debate is over the quality of the information. It’s not hard to find individual data series that will back any given argument. You have to look at a broad span of data.

对于信息质量的争论已经结束了。如果只看单个数据序列，很容易佐证任何的论点。一定要看比较广泛的数据。

Fred Hu, founding partner of Primavera Capital Group

胡祖六(Fred Hu)，春华资本集团(Primavera Capital Group)创始合伙人

They have over time really made efforts to improve the quality and reliability of economic data, but it’s obviously still a work in progress. I would say economic data, with a few exceptions, are released to the public on a regular basis.

长期以来，他们确实在努力提高经济数据的质量和可靠性，但很显然，这方面还需更多的努力。我想说，中国定期向公众发布了经济数据，只有少数例外。

My country still doesn’t permit freedom of speech and freedom of information. Most public discussions are severely restricted or penalized. Again, in the economic arena, it has been a notable exception.

我国仍然没有言论和信息自由。大多数公开讨论都受到了严格限制，甚至会遭受惩处。但经济领域却是一个明显的例外。

In social sciences, economists tend to have the biggest degree of freedom in discussing economic issues or policy issues, and economists can be highly critical of government policies without undue personal consequences.

在社会科学领域，经济学家在讨论经济或政策问题时，往往拥有最高的自由度。经济学家对政府的政策可能会有很强的批判性，而个人不会承担太大的后果。