Management and energy efficiency in a Chinese manufacturing cluster

Last registered on October 28, 2019


Trial Information

General Information

Management and energy efficiency in a Chinese manufacturing cluster
Initial registration date
October 27, 2019
Last updated
October 28, 2019, 11:17 AM EDT


Primary Investigator


Other Primary Investigator(s)

PI Affiliation

Additional Trial Information

Start date
End date
Secondary IDs
We investigate the effects of an energy management training designed to identify both technology and process-oriented energy-saving opportunities for firms in a machine components manufacturing cluster in Shandong Province, China. We aim to understand to what extent, and through what mechanisms, the intervention affects firms' energy management practices and ultimately improves their efficiency of energy use. We are interested in studying the relationship between measures of good management practices and the adoption of experts' recommendations. Our study will also allow us to qualitatively probe the role of management and other firm characteristics, such as firm size, degree of centralization, and appointment of a dedicated energy manager, on achievement of energy savings through a set of pre- and post-intervention survey questions. The training is conducted by a team of local and international energy saving experts with expertise in the metalworking equipment and production processes employed by the firms in our sample. For each firm, experts generate a customized list of improvements and deliver their findings to the firms in the form of a report. A representative of the team follows up every three months to assess progress and elaborate or adjust recommendations.
External Link(s)

Registration Citation

Karplus, Valerie Jean and Da Zhang. 2019. "Management and energy efficiency in a Chinese manufacturing cluster." AEA RCT Registry. October 28.
Sponsors & Partners

There are documents in this trial unavailable to the public. Use the button below to request access to this information.

Request Information
Experimental Details


From June to September 2017, one MIT representative, 1-2 McKinsey representative, and 1-2 representatives from the Shandong Zhengxiang International Low-Carbon Technology Company will administer a training aiming at improving firms' energy efficiency to a sample of 20-24 firms randomly selected from the total population of 48 surveyed firms. The training will last one to two days with regular follow-up every three months, with troubleshooting as needed. During the training, firms' current energy management and energy use situation will be assessed, and recommendations tailored to the firms will be provided. Each recommendation will be accompanied with a detailed assessment of energy and cost savings potential. There are in general three types of recommendations: energy management, technology upgrading, and policy information. Energy management-related recommendations will provide several key energy management best practices that firms have not yet adopted, e.g. establishing energy baselines, setting energy targets, appointing an energy management team, and providing them with effective incentives; Technology upgrading-related recommendations will provide information that can help firms to replace or upgrade energy equipment, e.g. identifying outdated air compressor or motor and providing upgrade suggestions, proposing a more efficient lighting design, and adjusting production activities to reduce idle time; Policy information-related recommendations will inform firms of relevant energy saving (either in terms of quantity or cost) policies that the central and local governments have already provided, e.g. direct power purchase that provides lower electricity tariff available for some firms, heating fee that is based on actual energy use not heating area, and energy-saving subsidies that firms may be qualified to apply when upgrading technology.

Firms will be made aware that researchers are piloting this new approach designed to identify these energy efficiency recommendations based on international experience and tailored to their individual circumstances. The recommendations will be provided in a written notice and an online software application, known as the RedE tool, developed by McKinsey.
Intervention Start Date
Intervention End Date

Primary Outcomes

Primary Outcomes (end points)
Recommendation adoption status, and
firm energy use and efficiency (monthly electricity use and electricity use per unit of output)
Primary Outcomes (explanation)
Management level: a variable that is derived from the scores of 18 questions in a survey that is developed to understand firms' management practices based on Bloom et al. (QJE, 2007), and
energy management level: a variable that is derived from the scores of 10 questions in a survey that is developed to understand firms' energy management practices based on the ISO 50001 and Chinese national standard GB 23331 for energy management.

Secondary Outcomes

Secondary Outcomes (end points)
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
In the baseline survey, we collected energy use (monthly electricity use) and output (in quantity and value term) of 100 firms in the last three years (2014-2016) as well as firm characteristics such as ownership, industry code, export status and etc. We also implemented a survey to measure the management and energy management levels before the intervention.

We choose 48 firms that satisfy the requirements of our study and randomly choose 24 firms to receive the intervention from June to September 2017. After the training is completed, we will provide follow-up support and track the adoption status of the recommendations of treatment firms for one and half years. For all the participating firms, we will collect post-treatment energy use and output information. By the end of 2018, we should be able to compare the energy efficiency and energy use as well as management and energy management levels between the treatment and control firms.
Experimental Design Details

After providing the recommendations to the firms, we plan to review the adoption of recommendations and firm energy use changes regularly. Two months after all the training is completed, we will let the line managers or persons who are our main contacts and able to adopt the recommendations in these firms know that they will be awarded if recommendations are adopted (each adoption for RMB 500, up to five adoptions). Among half of the treatment firms randomly selected, we will let them know that adopting any recommendation from the list that we provided to them will be qualified to receive the award; among the other half of the treatment firms, we will provide them a short list of recommendations that we think are the most effective recommendations and let them know that the awards are only for the adoption of the recommendations on the short list. By making the incentive more clear and information more salient, we are interested in finding whether the adoption and energy efficiency improvement later could be more significant.
Randomization Method
Using the pair-wise matching method described in Bruhn and McKenzie (AEJAE, 2009), we created 24 pairs of firms out of the 48 firms that participate in the experiment. We use firms' electricity consumption, electricity intensity, sales and sales growth in 2015 as well as management and energy management levels in the matching process. In each pair of the firms, we randomly choose one as the treatment firm and the other as the control firm.
Randomization Unit
Was the treatment clustered?

Experiment Characteristics

Sample size: planned number of clusters
48 firms
Sample size: planned number of observations
48 firms
Sample size (or number of clusters) by treatment arms
24 firms control and 24 firms treatment (12 in one treatment arm and 12 in the other)
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Using the pair-wise matching method described in Bruhn and McKenzie (AEJAE, 2009), we developed a simulation to test the power of the design. In 100 simulation runs with expected minimum detectable effect (8% electricity intensity reduction in quantity term), we were able to detect the treatment effect with a p-value smaller than 0.10 in 73 runs.

Institutional Review Boards (IRBs)

IRB Name
Committee on the Use of Humans as Experiments Subjects, Massachusetts Institute of Technology
IRB Approval Date
IRB Approval Number


Post Trial Information

Study Withdrawal

There are documents in this trial unavailable to the public. Use the button below to request access to this information.

Request Information


Is the intervention completed?
Data Collection Complete
Data Publication

Data Publication

Is public data available?

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials