Many leaders say they want to delegate more and then find themselves approving everything. A well-known field experiment suggests the problem is often not trust alone but information: people can only be given decisions when the facts they need are in front of them.
What the Indian textile experiment tested
Researchers Nicholas Bloom, Benn Eifert, Aprajit Mahajan, David McKenzie and John Roberts published the study as a National Bureau of Economic Research (NBER) working paper in January 2011. It is titled “Does Management Matter? Evidence from India.” A working paper is research circulated before formal journal publication, so it is best read as strong evidence rather than a final verdict.
The team ran a management field experiment on large textile firms in India. Some plants were randomly chosen to receive free consulting on modern management practices. Other plants served as a comparison group. Random selection matters: it makes it far more likely that differences between the two groups come from the consulting rather than from pre-existing differences between firms.
The practices themselves were ordinary operating discipline: measuring quality, tracking inventory, recording what happens on the production floor and using those records to make decisions. Nothing about them was glamorous, which is part of the point.
What changed: productivity and decision-making
According to the NBER abstract, adopting the practices raised average productivity by 11%, through improved quality and efficiency and reduced inventory. Productivity here means how much output a plant gets from the resources it uses.
The authors also report a second result that is easy to miss. Decision-making became more decentralized. In plain terms, owners delegated more decisions to middle managers, and the authors link this to better information flow. Computer use also increased, since collecting and analyzing the data required it.
The sequence is the useful part. The owners did not first decide to trust their managers more and then hope for the best. They first built a reliable way to see what was happening, and delegation followed.
Why would good information make delegation easier?
Consider what a leader needs before letting go of a decision. They need to know whether the decision is being made well, whether problems will surface early and whether the person deciding has the same facts the leader would use. Without shared data, the only safeguard is for the leader to review everything personally, which is exactly what keeps decisions centralized.
When information is recorded and visible, three things change:
- Quality can be checked after the fact. A leader can review outcomes rather than approve every step.
- Managers can act on the same facts. They no longer need to ask the owner what the numbers say.
- Mistakes become learnable. A bad call shows up in the data, and the lesson can be shared rather than hidden.
This is a mechanism, not a guarantee. The study documents the pattern in one industry and one country, and the abstract describes the link between information and delegation without promising the same result everywhere.
Why didn’t firms adopt these practices sooner?
If the practices were profitable, why were they not already in place? The authors ask this directly. Their conclusion, in the abstract’s words, is that “informational barriers were a primary factor in explaining this lack of adoption.” Many firms simply did not know the practices existed or what they could deliver.
They also note that limits on firm entry and growth meant poorly managed firms were not quickly driven out of the market. In a competitive market, weak management gets punished fast. Where competition is softer, weak habits can persist for years.
The authors sum up the finding in one line: “Modern management is a technology that diffuses slowly between firms.” For a team leader, the smaller-scale version is familiar. A better habit inside one department may not spread to the next office, not because anyone resists it but because nobody has seen it work.
A worked example: from approval queue to guardrails
The following is illustrative arithmetic, not data from the study. Imagine a regional operations lead whose team of eight must get approval for any customer refund. If each approval takes five minutes of the lead’s attention and the team raises 12 refunds a day, that is 60 minutes a day (12 × 5), or five hours a week over a five-day week.
Now suppose the lead publishes a simple weekly report showing refund counts, amounts and reasons, and sets a guardrail: team members may approve refunds under a stated limit, and anything above goes to the lead. If nine of those 12 daily refunds fall under the limit, the lead now reviews three a day, or 15 minutes (3 × 5). That frees 45 minutes a day, or 3.75 hours a week.
The saving is only part of the benefit. The team also responds faster to customers, and the lead’s review time shifts from rubber-stamping to watching patterns. The report is what makes the guardrail safe.
| Before | After |
|---|---|
| Lead approves every refund | Team approves under a set limit |
| Lead has no standing view of patterns | Weekly report shows counts, amounts, reasons |
| Delay while waiting for the lead | Same-day answers for most cases |
| Mistakes found by accident | Mistakes found in review of the report |
Which decisions are safe to hand over?
Not every decision should be delegated, and the experiment does not claim otherwise. A practical screen asks four questions:
- How costly is a wrong call? Small, reversible errors are good candidates. Large or irreversible ones usually are not.
- Does the decider have the facts? If not, fix the information first.
- Will you learn quickly if it goes wrong? Fast feedback makes delegation cheap to correct.
- Is the decision a developmental stretch? Handing over a well-bounded decision is one of the most direct ways to grow someone’s judgment.
The article Decision-Making Under Uncertainty: A Framework for Leaders offers a longer way to think about reversibility and risk, which pairs naturally with the screen above.
What does this mean for the manager’s own job?
Delegation changes what managers do rather than reducing it. Once routine approvals move to the team, the leader’s time goes to setting guardrails, maintaining the information that makes them safe, coaching and clearing obstacles. What Managers Actually Do: What the Research Says describes that broader set of responsibilities.
Two habits make the shift easier to sustain:
- Review outcomes on a rhythm. A weekly or monthly look at the numbers beats ad hoc spot checks and keeps the leader informed without hovering.
- Give feedback that is specific. When a delegated decision goes poorly, the response shapes whether people take future decisions seriously. How to Give Feedback That People Can Actually Use covers how to do this without discouraging initiative.
Choosing the right measures is its own skill. What Output Per Hour Tells Your Team explains how productivity measures can inform a team and where they mislead.
Limits of the evidence
A few cautions apply before borrowing the lesson:
- Context. The study covers large textile firms in one country. A software team, a clinic and a factory may need very different information to delegate safely.
- Preliminary status. It is a working paper, and the abstract summarizes results; the full paper contains details this article does not repeat.
- Correlation inside the result. The abstract links better information to more delegation, but a leader should not assume that every dashboard produces delegation or that every delegation produces gains.
- Measurement risk. Data can be wrong, narrow or gamed. Measuring only what is easy can steer a team toward the wrong target.
Start with visibility, then loosen the grip
The core lesson of the experiment is about order of operations. Better routines for recording and sharing information raised productivity by 11% in the studied plants, and they let owners hand more decisions to middle managers. Leaders who want to delegate more can begin the same way: pick one recurring decision, make the relevant facts visible, set a clear guardrail and review the results on a schedule. Trust then rests on evidence rather than hope.
This article is general information about management practice and research, not legal, employment or business advice.
Sources: National Bureau of Economic Research
